bitshares-consensus.tex

\documentclass{btswhitepaper}
\title{BitShares 2.0: Consensus Mechanism}
\input{AUTHORS}
\begin{document}
\maketitle

\begin{abstract}%
\end{abstract}
\section    { Introduction                                   } \input { content/cons          } 
\section    { Delegated Proof-of-Stake (DPOS)                } \input { content/cons-dpos     } 
\section    { Transactions as Proof-of-Stake (TaPOS)         } \input { content/cons-tapos    } 
\section    { Distinction from Traditional Consensus Schemes } \input { content/cons-comp     } 
\section    { Attack Vectors                                 } \input { content/cons-att      } 
\section    { Conclusion                                     } \input { content/cons-conc     } 
\section*   { Litarture                                      }
\nocite{*}
\bibliographystyle{IEEEtran}
\bibliography{literature}
\end{document}

% Randomness of Witness Scheduling
% https://bitsharestalk.org/index.php/topic,18126.msg231771.html#msg231771


% https://bitsharestalk.org/index.php/topic,18613.msg239949.html#msg239949
% Compromising one witness does not give you unlimited ability to perform a
% double spend attack.   To perform the ``double-spend'' they would have to
% broadcast the transaction 1 block before their turn, then skip the block that
% included their first transaction and produce a block that contained an
% alternative transaction\ldots *AND* have cooperation of the next witness which
% would have to be in on it because that next witness would have seen the
% original block/transaction first and would therefore ignore the bad witnesses'
% block.
% 
% This means that you need at least 2 witnesses to attempt a double spend and you
% could only do it any time those two witnesses were randomly selected to follow
% one another.   
% 
% Lastly you could only perform the double spend attack as part of an anonymous
% transaction (ie: paying someone who does not know your identity) and the
% transaction and the other party would have to accept single confirmation
% transactions and then take some kind of irreversible action.   As a result
% meta-exchange, block trades, and exchanges would require several witnesses to
% sign (up to 51% of the witnesses).    Because most witnesses will be publicly
% known with real identities and reputations on the line, the risk of criminal
% charges for a provable intentional double spend that is not rectified is enough
% to prevent it from happening altogether.
% 
% The opportunity to execute large, anonymous transactions involving irreversible
% actions in less than 10 seconds will be vanishingly small.   Once it was
% detected everyone would be on notice to wait 10 seconds until 10 of 19
% witnesses have signed and to vote out the attackers.    Therefore, you could
% probably pull off the double spend ``ONCE'', the profit earned would be small
% compared to the value of the income from the witnesses.


% https://bitsharestalk.org/index.php?topic=20210.msg255901;topicseen#new