Item — HN Reader

I think about this a lot. There's the obvious archival problem, to ensure the data is preserved along with a machine to read it or instructions on how to build such a machine.

But then there's the retrieval problem. I'm very interested in family history and genealogy. Every journal, every photo from even a couple generations back is like a priceless treasure. 100 years from now? Our great-grandkids' grandkids will have more data on us than they'll know what to do with.

One of my long term projects I'd like to help tackle is to answer the question "How do you summarize a life?". I like to imagine a system similar to Superman's ice castle where you insert a digital artifact from your ancestor and an AI helps you sift through their data.

1 comments

> to ensure the data is preserved along with a machine to read it or instructions on how to build such a machine

You don't need to preserve the machine. You just need to be able to emulate it sufficiently close on then-current hardware. The nice thing about this approach is that if it's done right, then it scales more or less indefinitely, since you can run emulators inside other emulators. So long as at any given point we have one that covers the previous generation...

This takes care of file formats and software. Maintaining compatibility with the physical medium on which data is archived is trickier, because you can't emulate the reader. But on the other hand, with digital data, you can just copy it over to the new medium. So, again, as long as the archives are regularly updated to keep up with the tech - not even current, just whatever is still available - we're good.