[Bitcoin-development] Bitcoin TX fill-or-kill deterministic behavior

[Jeff Garzik]

On Sat, Apr 14, 2012 at 11:13 AM, Mike Hearn <mike at plan99.net> wrote:
>> So, to be specific... a A->B chain of transactions, that collectively
>> meet the network's fee requirements?
>
> Yes.

ACK on the concept

>> Ideally the fee, if any, is market based and negotiated. Problem is... like
>> democracy, no matter how ugly it is, people have trouble finding a
>> better system :)
>
> I think this is something we can explore over the coming years. I
> favor having people commonly pass transactions around outside the
> broadcast network with the transactions and their dependencies being
> broadcast only when there's a lack of trust between recipient and
> sender. The block chain is an optional service after all.

Agreed.  A TX is just a signed message.  No reason why it -must- use
mainnet's distributed notary service.

>> Furthermore, many of these ideas -- like sending TX's directly to the
>> merchant -- involve far more direct payee<->payer communication on the
>> part of the wallet client than is currently envisioned
>
> Yes, though it's worth remembering that the original Bitcoin design
> did have participants communicate directly. When I talked with Satoshi
> in 2009 he saw the pay-to-IP-address mode imagined as the normal way
> to make payments, with pay-to-address being used as a kind of backup
> for when the recipient was offline.
>
> In the end that's not how things evolved, but it the pendulum could
> easily swing back the other way.

IIRC pay-to-IP was removed because it was unreliable -and- detrimental
to privacy?  ISTR Satoshi specifically disliking the privacy elements
of p2ip.

But I also have a "gut feeling" that these sorts of payments and
direct communication should be done via a wholly separate protocol
than the bitcoin P2P protocol.  Doing p2ip as it was done originally,
inside the bitcoin P2P protocol, was a mistake.  Extensible as it is,
I think a better job -- and faster evolution -- can be done with a
separate protocol on a separate port.

Some HTTP derivative would probably make life easier for mobile
payments and firewalled scenarios, and for client->merchant
communications, for instance.

-- 
Jeff Garzik
exMULTI, Inc.
jgarzik at exmulti.com