The Wayback Machine is fussy about what types of file it archives. I can’t find a relevant FAQ, but it appears to only archive text (HTML) and images. For example, this little-known amazingly unpopular and unlinked image is archived, whereas putty.exe, a well-known super-popular program with high pagerank is not.

Based on this evidence, I’m assuming that uploading my multi-gig backup files to the web won’t impress Archive.org too much. So:

Here’s an experiment.

The file 20th c.gpg contains a text file containing my credit card number, expiry date and CCV code, and a TeX file containing a proof of the Riemann Hypothesis.

20th c.gpg.asc is a plain-text ASCII-armored version of the same file created using this command: gpg --enarmor 20th\ c.gpg

20th c.gpg.txt is the same as above, but with a different filename extension.

20th c.gpg.jpg is the first file embedded in a picture using this command: steghide embed -cf 20th\ c.gpg.jpg -ef 20th\ c.gpg -p 123

20th c.gpg.embedded.jpg is a copy of the same file, but embedded in this page here:

I predict that this last image will definitely get archived, and possibly the plain text versions too. I’m also willing to bet that the NSA automatically archives every .gpg file it can get its hands on, so it can put my name on its list of troublemakers. It probably wants that proof too.

N.B. This doesn’t appear to be against Archive.org’s Terms and Conditions, because publishing a website doesn’t make me a user of their services. But I’m sure they will remain happy to manually remove my site from the Wayback Machine if the experiment is successful and I start inserting massive quantities of encrypted private data into their system. Plan B: package all my data into a series of Ogg Theora files with only the bare minimum formatting required to produce a valid Theora bitstream, name it “DocumentaryOnExperimentalElectronica.ogg”, and upload it to the movie archive! Perfect evil.