Suggestion: Payment Return Addresses

[NewLibertyStandard]

I suggest Bitcoin optionally be able to include a Bitcoin return address when sending a payment. Right now if someone sends me too many bitcoins I can't send them change without asking for a return address over email which slows down the transaction considerably. If they could include a return address, then I could immediately send them their change. Of course it should still be possible to not include a return address for when people do not want the option of change or a refund. If a person wanted to donate their pocket change to a charity, they could set their return address to that of a charity and round their payments up to the next whole bitcoin.

On the downside people could abuse the feature by including a text message within a fake Bitcoin address. Then everyone's hard drives might become more filled if people were using transactions as instant messages. But I doubt people would use Bitcoin as an instant messenger since messages take ten minutes to send and would be public to whoever was so bored (and technically capable) as to read the bitcoin addresses being sent. Bitcoin could be made to only allow addresses to which Bitcoin has the private key, but that would disable the donate your pocket change option I mentioned. And of course the client could be modified to not require the private key of the return address.

Some of these disadvantages actually already exist in Bitcoin. If someone wanted to hide a message to the public, they could send ฿0.01 to a fake bitcoin address which includes their message. And if someone wanted to fill up other people's hard drives, they could write a script to repeatedly send bitcoins back and forth between two clients. Hmm, I think it would be advisable to come up with a solution to that second weakness of Bitcoin regardless of whether return addresses are implemented.

I think that the advantages of this feature would outweigh the possible disadvantages. What do you think?

[1715292105]

1715292105

[1715292105]

1715292105

[1715292105]

1715292105

[BitcoinFX]

Good idea. I also think that the advantages would outweigh the disadvantages. It would need to be 'optional' for sure.

However, from my basic understanding of public key cryptography I can't conceive a method to implement this without affecting anonymity. Moreover, as the 'service provider' you are already in contact with your customer. Therefore, why 'leak' some of this information with other bitcoin nodes.

It would be a bit like signing a 'mixminion' email with the actual 'from' address. 

[NewLibertyStandard]

There is nothing that ties a bitcoin address to your identity other than that you can unlock bitcoins sent to it. You can use a separate address for each transaction, each person who might send you bitcoins or you can use the same address for all your transactions.

[tdahsu]

I'm curious - genuinely - in what situation would there be "change" to give back?  Why would you not specify a price that doesn't involve it?

[frozen]

Why not just include a return address in the public note?

[Trongersoll]

I'm curious - genuinely - in what situation would there be "change" to give back?  Why would you not specify a price that doesn't involve it?

Fat Fingers?

[Sukrim]

Why not just use one of the input addresses that they sent the BTC from?

Hosted wallets...

And a possible useful feature would be to charge slightly more for something in BTC, do a market trade on an exchange to get the USD/EUR/WTF you want for whatever you're selling and return the remaining BTC along with the item. No need for Bitpay that way.

[kiko]

Please do keep up people.

Gavin and co. are working on this as a formal protocol.

read this: https://gist.github.com/gavinandresen/4120476

[agaumoney]

I'm curious - genuinely - in what situation would there be "change" to give back?  Why would you not specify a price that doesn't involve it?

Actual shipping was less than the amount billed at the time the order was placed.

[kjj]

Wow.  37 month thread necro.  New record?

[bbit]

Wow.  37 month thread necro.  New record?

[tdahsu]

Ja, I saw that when I posted my question... the thread even warned me that no one had posted in a really long time.  Of course I did not let that stop me!  I *was* genuinely curious after all!   

[CasinoBit]

You can pull out the "return" address pretty easily with a series of hashing alogrithms, it's nice because it naturally prevents newbies from creating applications that send back automatically similar to how we work.