4

For example, how can a wallet determine what addresses correspond to a given seed (before and after a snapshot)?

new_xxx
  • 105
  • 4

1 Answers1

5

An address is generated from a private key.

A private key is generated from any random 81 trytes (e.g. a seed)

If you want to know how this generating works in detail, you should look at the implementations in the GitHub. (e.g. here in PyOTA) It's basically just a lot of Kerl/Curl hashing.

Since we want to create more than 1 private key from a seed, we do this:

  • private key 0 is generated from seed
  • private key 1 is generated from seed + 1
  • private key 2 is generated from seed + 2
  • private key 3 is generated from seed + 3
  • ...

Since the seed consists of trytes, it's basically just a number in ternary format, so we can add to it.

If you already used the first 10 addresses of your seed, you could theoretically use seed + 10 as your new seed and you would still see your total balance in a wallet.

I made a program that does exactly that. It trims your seed so that your used addresses aren't generated anymore when you log in to a wallet.

(Here is the part where I add a number to a seed.)

Zauz
  • 4,454
  • 15
  • 42
  • In a wallet, should there be the ability to generate a new address at a user's request or should a new address be generated automatically under certain circumstances (e.g., after an old address has been used as the input of some transaction)?
    • – new_xxx Feb 15 '18 at 18:12
  • Should a wallet store or in some way keep track of all the addresses it has generated? Also, should it mostly ignore the “used” addresses and use only the “active” one(s) in a meaningful way?
    • – new_xxx Feb 15 '18 at 18:12
  • And why do they say that addresses should be regenerated in a wallet after a snapshot? Isn’t it so that a wallet should constantly have all the required info about its addresses, why does a snapshot require it to take some extra steps to reobtain such info?
    • – new_xxx Feb 15 '18 at 18:12