bitcoind transaction to ip address

5 messages BitcoinTalk lfm, Satoshi Nakamoto, knightmb, Laszlo Hanyecz August 5, 2010 — August 6, 2010
lfm August 5, 2010 14:22 UTC Source ·

I cant figure out how to send a transaction to an ip address from bitcoind command line interface. Has the function been implemented yet? (linux 64 if it matters)

Satoshi Nakamoto August 5, 2010 17:28 UTC Source ·

It’s not implemented.

It turned out nobody liked that mode of transfer anyway, so it hasn’t had much development attention.

knightmb August 5, 2010 18:17 UTC Source ·
Quote from: satoshi on August 05, 2010, 5:28:40 PM UTC

It’s not implemented.

It turned out nobody liked that mode of transfer anyway, so it hasn’t had much development attention.

Well I guess it’s time it got some attention 😉

Dang, that’s two things I have to put my list now, good thing it’s open source! 😄

Laszlo Hanyecz (laszlo) August 6, 2010 01:19 UTC Source ·

If someone takes the ip address method to the next level it really should be changed to provide some kind of security, like agree on a password/address or something ahead of time. Sending to an address and creating the key automatically provides no authentication and there’s no recourse if it went to the wrong place; the intended receiver is unaware you even tried to send anything. It is trivial to proxy it and MITM that kind of activity and with the value of bitcoins rising it might just be worth doing eventually.

I consider the ip address sending a local testing feature more than a production level feature in the way it is currently implemented. One possibility that could work is if the receiver provided a public key (address) to you first and you used a combination of internet address and public key. It is not as convenient but the alternative is to just send it to the ether and hope nobody grabs it before it gets to your target. If anyone has used liberty reserve or something similar you know it is certainly not convenient to log into and such, but when it comes down to a choice of losing your money or dealing with an inconvenience I think most people would prefer the latter.

knightmb August 6, 2010 03:11 UTC Source ·
Quote from: Laszlo Hanyecz on August 06, 2010, 1:19:12 AM UTC

If someone takes the ip address method to the next level it really should be changed to provide some kind of security, like agree on a password/address or something ahead of time. Sending to an address and creating the key automatically provides no authentication and there’s no recourse if it went to the wrong place; the intended receiver is unaware you even tried to send anything. It is trivial to proxy it and MITM that kind of activity and with the value of bitcoins rising it might just be worth doing eventually.

What if it had a flag for “secure/unsecure” transaction type, where one can have a password for static IP to static IP transactions and an unsecure mode that works like the default IP sending the GUI uses?

I consider the ip address sending a local testing feature more than a production level feature in the way it is currently implemented. One possibility that could work is if the receiver provided a public key (address) to you first and you used a combination of internet address and public key. It is not as convenient but the alternative is to just send it to the ether and hope nobody grabs it before it gets to your target. If anyone has used liberty reserve or something similar you know it is certainly not convenient to log into and such, but when it comes down to a choice of losing your money or dealing with an inconvenience I think most people would prefer the latter.

If the address based transactions types could send the comments/from field like the direct IP to IP, that would be another reason not to really use the feature (except for testing) 😉