Een papieren internet

If you wanted to preserve important bits of our civilization for future centuries, you could do worse than a bundle of paper sealed in plastic. It's remarkably cheap and effective; you can make one over a weekend.
(...)
A time capsule must perform three basic functions:

1. Encode information with sufficient density & durability.
2. Protect the information from physical damage, moisture, heat & cold, etc.
3. Be findable.

So while the Rosetta Stone performed (2) fairly well, it was pretty lucky to be found at all. Also its data density is terrible: about 1 bit per cubic centimeter. A book in a library fulfills (1), but requires the library around it to provide (2) and (3).
The internet, contrary to popular belief, is not very good at preserving information on a long time scale. It ultimately depends on digital media that break down rapidly. Early Unix source code, one of the most important sequences of bits ever written, had to be reconstructed from printouts.

Het printen van internet, ik geloof toch niet dat dit de ultieme oplossing is...

Tegelijkertijd publiceren een Duitser en Oostenrijker een onderzoek waaruit blijkt dat opslag op microfilm de ideale oplossing zou zijn voor de digitale houdbaarheid.

The team has carried out a feasibility study that analysed encoding techniques to allow digital data to be saved on to microfilm and then to test data recovery as well as cost issues. Aside from precluding the need for frequent technology updates, storage of documents and data on microfilm will give future generations access to the information by scanning the microfilm into whatever system they are currently using and applying optical character recognition to re-digitize und subsequently decode the data.
The team further suggests that in order to reduce the amount of microfilm used for any given repository and so cut conversion and re-digitization times it would be possible to convert a stream of text into a bar-code type system that would still be entirely analogue but would rely on knowledge of the conversion key to return the data to digital form from microfilm. Using such a system could render a tested 170 kilobyte file that requires 191 pages of microfilm space as just 12 or so "printed as a two-dimensional barcode. Such a barcode would incorporate redundancy and be self-checking unlike a straight digital to analogue image scan of the text. Further compression is possible, if colour microfilm and barcodes were used for storage. This may provide a valuable, low-maintenance additional back-up for the original digital objects in addition to preservation activities needed for the on-line access copies.