Very interesting blog indeed. I totally agree, I think this is the ‘double edge’ knife of abstraction, both a source of advantage yet a source of failure too.

The only thing I would comment on regarding the analogy; is that it is slightly flawed in the sense that XOR gate would I assume eventually feed into a Schmitt trigger hence I doubt one would see those actual voltages!

But I guess that just further validates your point :) regarding abstraction and underlying implementation.

Did you make this a recursive analogy on purpose :0)

A perfect encryption system generates an encoded data stream that has perfect entropy, that is, any given symbol (bit or byte or other) has the same probability of occurrence as any other symbol. The probability is 1/N where N is the number of symbols. If you have this condition, there is no information discernible to the man-in-the-middle from the data stream itself.

Does the encryption actually have all the same characteristics as maximum entropy? Usually it will not. Any one difference is the pry bar that cracks codes.

In this case, the analog voltage levels are additional non-maximal entropy information order that is overlaid over the otherwise random data stream. The 2 logic threshold are an additional 1 bit of information that is modulated to the plain text directly and simply added atop the data stream.

Since digital is nothing more than a useful human-created abstraction – everything is actually analog – there is always a flaw in the armor somewhere.