[Bitcoin-ml] Malleability Fix SIGHASH_ANYOUTPUT

[Steve]

> In my mind, everyone should be able to use & innovate Bitcoin in as
> many unique ways as they want to so long as their innovation doesn't
> actively limit or harm some other uses, applications, or security of
> Bitcoin in a measurable way.

I greatly prefer this proposal to malfix and FT for exactly the reason 
that it doesn't alter any existing ways of using Bitcoin and yet adds 
the functionality that is most often cited as the reason for a malfix.  
Although I'm not quite sure who is actually asking for 0-conf 
transaction chains.

> I briefly proposed and dismissed this as it is a very dangerous flag due
> to address reuse: The recipient of such transaction can "replay" it for
> every output with the same address and amount.

Including the output index in the signature data doesn't solve this 
problem but it does by default reduce the set of outputs that are 
repayable and potentially opens up some other clever uses.  As it 
doesn't affect the stated goal in any way I can so no reason not to 
include it.

Gmax had an interesting suggestion in response to the proposed 
SIGHASH_NOINPUT which has the same replayability issue.  He suggested 
naming the flag SIGHASH_REPLAY_VULNERABLE.  I might suggest 
SIGHASH_REPLAYABLE.  As a useful means of making sure people who use it 
have the issue called to their attention.
https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2016-February/012461.html