Bijective Mnemonic Phrases for addresses, hashes, etc


I hope this is not too off-topic.
When doing data analysis on the blockchain, I have to deal with a lot of addresses. Trying to remember and discern them gives me headaches. I tried to offer a solution here:

I need your feedback on this.

There is also an issue with my proposal: the first word of most mnemonics start with the letter ‘a’, which kind of defeats the purpose. This is not due to an error, 20 byte addresses fall into a range where the base conversion results in this accumulation for the first word. I have a few solutions in mind, but first would like to hear your opinion.


I now realize this is off-topic. I couldn’t delete the post, so you can do it if you’re a mod


I quite like it actually! I really think a different set of words than BIP39 would be required. I cringe thinking about someone inadvertently copy/pasting their seed. I know you design it against that, but it WILL happen lol. At least an input field can reject the word list as invalid (although sizes do the same thing)

You proposal has checksums, right?


If you do three words (like status does when verifying Transactions) you could probably stick that checksum word at the end, and it would enough entropy to be hard to spoof

Also, imagine if you check summed with colors?


Oh right, we do need to use a different word set if there is to be an app that takes phrases as input.

My main purpose was to increase readability. If people use this as a way of sharing public keys, e.g. on the phone, we have to make sure they don’t share seeds. I’m open to suggestions.

The proposal doesn’t have checksums. Do you mean computing the checksum of an entire phrase, and appending it at the end of its truncated version? Sounds interesting, is this to decrease the risk of collision?

Wallet Shape Address Human-friendly Visualization of Wallet Address : Momcode

Decreases the risk of collision, but first and foremost protects against the user making a mistake flopping a word or using the wrong one. Capital and lowercase letters in hex addresses are used to to track the checksum of the address for this purpose.


By shuffling the word list I get slightly better visual checksumming:


@Cygnusfear Good idea, and simple too.

Still, the first word will be one of the first 100 from the list most of the time. I should try to find a bijective map that

  • addresses this problem,
  • keeps the same number of words/digits per hexadecimal number,
  • is easy to compute.

Going to delve into the issue when I find the time.