>It fits into my thinking about the Long Now. Working in formats that can be opened in 100 or 1000 years. This is a real problem for old stuff from decades ago

>It fits into my thinking about the Long Now. Working in formats that can be opened in 100 or 1000 years. This is a real problem for old stuff from decades ago, but much less so now: as long as the data is in some format that uses open-source code to read (i.e., open-source A/V codecs), you just have to make sure the source code is available for it. A lot of data from ages ago is really hard to decipher because it was stored in proprietary formats, and the specifications and code for those are lost or difficult to find. Data stored as PDF or h.264, for instance, probably won't have this problem.