Test Data: Visual Fingerprints

[tomrittervg]

We should extract and adapt the OpenSSH Visual Host Keys code to generate fingerprints. It should also generate two fingerprints that are matches to 2^80 complexity.

• One 'match' should have the difference completely at random
• One 'match' should have the location of the characters as close as possible, but the less 'computation' expended on the type of character

[trevp]

In addition to OpenSSH fingerprints we should consider "vash" and "Hash Visualization":

https://github.com/thevash/vash https://moderncrypto.org/mail-archive/messaging/2014/000217.html

Hash Visualization: https://sparrow.ece.cmu.edu/group/pub/old-pubs/validation.pdf

The OpenSSH algorithm is described here: http://aarontoponce.org/drunken_bishop.pdf

Figuring out which is best, and then how to simulate a 2^80 attack seems like a ton of work, probably a study / research project in itself. Dunno if we have the resources for this, or maybe we should scale back and just focus on the text formats?