I have a proposal for alternative Mastercoin encoding ready and I wanted to discuss the feasibility of my solution. We are going to use a OP_CHECKMULTISIG script with three (or more) public keys to encode the data.
Because we will be using public keys to encapsulate the data we now have 64 bytes to encode data (65 minus one byte for the 04 marker). In order to not polute the blockchain we will always supply a publickey that the sender owns as first signature this way the output can always be redeemed. The other signatures we supply will be data-addresses; these transactions won't be able to be converted to addresses since they are not technically ECDSA x/y coordinates but they should be relayed and mined just fine.
I've created a
test transaction on testnet3 (
raw) to test this functionality.
miKnddGDQfU6rRYpLp2dhRyttnxH1WUo21 is the exodus that I use on Testnet. I'm sending the coins to myself (miDyBpL3TeeJRbtwdFUsw9mUWUEhN6FyXf) in this example so I send both the change amount and 0.00006 transaction to this address. I also use this address as the first signature for the multisg output so I can redeem this later. If you check the raw transaction you will see that there are two more keys supplied. If you decode these using the encoding standard from the whitepaper (but instead pad with zeros until you have 65 bytes) you will see that I'm doing two transaction for 1 test Mastercoin each.
In theory you could do multiple Mastercoin transactions in one multisig output but I would like to propose to keep it simple and only one transaction per output. If you want to send multiple Mastercoin transactions in one Bitcoin transaction you can create multiple multisig outputs.
Since all current ideas of the whitepaper fit within 126 bytes (2 signatures) and signatures are in fixed order we might not need to add sequences to all signatures but for now let's continue doing it.
I know there are a lot of users with more knowledge of Bitcoin lurking so please let me know if this approach would work within the rules set out by the reference implementation. I'm especially interested in if we can use 1 out of 5 multisig transactions. I tried to make one; and it relayed; but I'm not sure it's actually a valid transaction type.
Please let me know your thoughts so we can continue building cool things on top of Mastercoin
![Smiley](https://bitcointalk.org/Smileys/default/smiley.gif)
Man, that was fast and bold. Big respect!