I changed the repo name for my monero shell stuff. If you happen to have cloned it you'll need to update the remote to point at the new repo. I just figure now that wallet generation is out of the way and I have all these functions I might be able to do other things (like sign messages) so calling the whole repo "wallet generator" may not make sense in the future.
https://code.zxcvbn.space/thatguyoverthere/monero.sh
I also tagged a "stable" release yesterday. It still runs into some unrecoverable errors sometimes, but it seems like it either fails or succeeds. The errors I have seen tend to be that it failed to generate a full mnemonic due to poor entropy or something similar. I don't believe it will give you a bad wallet address or mnemonics that won't give you private keys for said address. I've tested several wallets created this way now with both sending and receiving. I did run into a case where two mnemonics I generated actually shared a private key, but they both worked in wallet software so it wasn't my code that made that possible. In that situation I was testing the offset capabilities. Basically what happened was "mnemonic + offset = new mnemonic" but when reversing the process a different mnemonic was recovered than the original which still produced the same public wallet address.
If anyone here feels so inclined, having some extra testing and eyes on the code might uncover ways to limit the failures. This started as a learning exercise more than anything, so mission accomplished so far, but if it can result in useful tooling for the community also I'm game.
https://code.zxcvbn.space/thatguyoverthere/monero.sh
I also tagged a "stable" release yesterday. It still runs into some unrecoverable errors sometimes, but it seems like it either fails or succeeds. The errors I have seen tend to be that it failed to generate a full mnemonic due to poor entropy or something similar. I don't believe it will give you a bad wallet address or mnemonics that won't give you private keys for said address. I've tested several wallets created this way now with both sending and receiving. I did run into a case where two mnemonics I generated actually shared a private key, but they both worked in wallet software so it wasn't my code that made that possible. In that situation I was testing the offset capabilities. Basically what happened was "mnemonic + offset = new mnemonic" but when reversing the process a different mnemonic was recovered than the original which still produced the same public wallet address.
If anyone here feels so inclined, having some extra testing and eyes on the code might uncover ways to limit the failures. This started as a learning exercise more than anything, so mission accomplished so far, but if it can result in useful tooling for the community also I'm game.

