So yesterday I wrote a blog post about Bitcoin Address signed messages and how they might sometime, maybe be useful. Shortly thereafter I was thrust into a panic as I found myself awash in a sea of my own ignorance. I also must imagine I looked really stupid on IRC.
After a relaxing walk and reading I discovered that the ways Bitcoin Address signed messages were implemented in software as well as their verification was even more fragmented than I could have imagined. There seems to be nothing broken in the mathematics, but the code used doesn't seem to agree on the mathematics which should be used across implementations.1 And so I n00b n00b n00bed my way through learning things in the most effective way possible, through a painful public pile of fail.
These three most recent posts are brought to you by the Dunning-Kruger effect.
- a rather worrying though considering Bitcoin transactions are basically just signed messages [↩]