diff --git a/paper/paper.tex b/paper/paper.tex
index 63fa2b850f6ad11ee1b2f4a2ccf5e5afbc7bdda7..f32e43c98384ae7f34c6cfe074d9f7becc0c6776 100644
--- a/paper/paper.tex
+++ b/paper/paper.tex
@@ -159,7 +159,7 @@ Each block in the blockchain is made up of the block header (refer to figure \re
   }
 \end{lstlisting}
 
-The block header contains the version of the blockchain for which this block was created (\texttt{nVersion}) as a four byte field. The next field is 32 bytes long and contains the SHA-256 hash of the previous block's header (\texttt{hashPrevBlock}), not the hash of the entire block. This establishes the chain of answers we discussed previously. The next field is again a SHA-256 hash (\texttt{hashMerkleRoot}), but over the block data (i.e. all transactions of this block), which creates the connection between the block header and its content. Then there is another four byte field storing the timestamp of the block creation (in Unix time). Its value should be considered more as a hint and less as a precise date. The fifth field (\texttt{nBits}) describes the target difficulty of finding a block at the time the block was created. Lastly, the \texttt{nNonce} field is used as a nonce to set arbitrary values and therefore allow finding a block with the required difficulty.
+The block header contains the version of the blockchain for which this block was created (\texttt{nVersion}) as a four byte field. The next field is 32 bytes long and contains the \ac{SHA-256} hash of the previous block's header (\texttt{hashPrevBlock}), not the hash of the entire block. This establishes the chain of answers we discussed previously. The next field is again a \ac{SHA-256} hash (\texttt{hashMerkleRoot}), but over the block data (i.e. all transactions of this block), which creates the connection between the block header and its content. Then there is another four byte field storing the timestamp of the block creation (in Unix time). Its value should be considered more as a hint and less as a precise date. The fifth field (\texttt{nBits}) describes the target difficulty of finding a block at the time the block was created. Lastly, the \texttt{nNonce} field is used as a nonce to set arbitrary values and therefore allow finding a block with the required difficulty.
 
 When a miner sets the difficulty too low, hence increasing his chances of finding a block, other nodes simply reject the new block, as the difficulty of the new block does not meet their difficulty threshold. By this mechanism ``the network'' adjusts the difficulty automatically.
 
@@ -178,9 +178,9 @@ The basic procedure the miners follow is described by the code in listing \ref{l
   // publish blocks to other nodes
 \end{lstlisting}
 
-First, a new block header is allocated and filled with the relevant data (previous block hash, timestamp, etc.). Then SHA-256 squared hash of the block header is calculated. If this hash fulfills the required difficulty, we consider ourselves very lucky and publish the block to other nodes. These verify the solution, i.e. calculate the SHA-256 squared hash of the block header, and after successful verification publish the block to other nodes.
+First, a new block header is allocated and filled with the relevant data (previous block hash, timestamp, etc.). Then \ac{SHA-256} squared hash of the block header is calculated. If this hash fulfills the required difficulty, we consider ourselves very lucky and publish the block to other nodes. These verify the solution, i.e. calculate the \ac{SHA-256} squared hash of the block header, and after successful verification publish the block to other nodes.
 
-Bitcoin uses \emph{SHA-256 function squared}, due to the birthday attacks on the smaller but related SHA-1 hash. SHA-1's resistance to birthday attacks has been known to be vulnerable and a collaboration between Google Research and CWI Amsterdam has shown it is possible two create two distinct PDF files which result in the same SHA-1 hash \cite{SHAttered}.
+Bitcoin uses \emph{\ac{SHA-256} function squared}, due to the birthday attacks on the smaller but related SHA-1 hash. SHA-1's resistance to birthday attacks has been known to be vulnerable and a collaboration between Google Research and CWI Amsterdam has shown it is possible two create two distinct PDF files which result in the same \ac{SHA-1} hash \cite{SHAttered}.
 
 \section{Incentives}
 \label{sec:incentives}
@@ -201,7 +201,7 @@ Examining the expenses of Bitcoin mining on the other hand, we first take a look
 \section{Mining hardware}
 \label{sec:mininghardware}
 
-When Bitcoin first started in 2009 it was easily possible to mine blocks with regular CPUs and GPUs. However, as the power of the entire network increased, their hash rate was too low (i.e. not producing enough hashes per second), and they were replaced with FPGAs (Field-Programmable Gate Arrays). Nowadays it is only profitable to mine Bitcoin with specialized hardware, so called ASICs (Application-Specific Integrated Circuits).
+When Bitcoin first started in 2009 it was easily possible to mine blocks with regular \ac{CPU}s and \ac{GPU}. However, as the power of the entire network increased, their hash rate was too low (i.e. not producing enough hashes per second), and they were replaced with \ac{FPGA}. Nowadays it is only profitable to mine Bitcoin with specialized hardware, so called \ac{ASIC}.
 
 \begin{table}[hbt!]
   \centering
@@ -335,11 +335,11 @@ The table includes estimates from Digiconomist \cite{DigiconomistEU}, who used a
 \section{Ethereum's energy demand}
 \label{sec:ethereumenergy}
 
-Ethereum is another very popular blockchain. Unlike Bitcoin, its Proof Of Work algorithm is \emph{ASIC-resistant}, thus mainly GPUs are utilized for mining, as they can be reconfigured. This leads to a very diverse, decentralized and heterogenous landscape of miners all around the world, because anyone with a GPU can start mining Ethereum without the need of buying additional hardware etc. but it also leads to a higher overhead and generally less efficient mining.
+Ethereum is another very popular blockchain. Unlike Bitcoin, its Proof Of Work algorithm is \emph{\ac{ASIC}-resistant}, thus mainly \ac{GPU}s are utilized for mining, as they can be reconfigured. This leads to a very diverse, decentralized and heterogenous landscape of miners all around the world, because anyone with a \ac{GPU} can start mining Ethereum without the need of buying additional hardware etc. but it also leads to a higher overhead and generally less efficient mining.
 
 % TODO explain more here ?
 
-The NVIDIA GTX 1070 is currently the most efficient GPU for Ethereum mining \cite{HR1070}, with a conversion rate of $5$ megahashes per second (the Antminer S9 we mentioned for Bitcoin mining is at $13.5$ gigahashes per second, see section \ref{sec:mininghardware}).
+The NVIDIA GTX 1070 is currently the most efficient \ac{GPU} for Ethereum mining \cite{HR1070}, with a conversion rate of $5$ megahashes per second (the Antminer S9 we mentioned for Bitcoin mining is at $13.5$ gigahashes per second, see section \ref{sec:mininghardware}).
 
 Just like we did for Bitcoin, the hash rate of the entire Ethereum network can be derived from the block difficulty. Ethereum is currently at $125$ terahashes per second (from \url{etherscan.io}, accessed 2017-12-14 \cite{EtherscanHashRate}).
 %  Hashrate: $H_R = 125$ TH/s (from \url{etherscan.io} %\autocite{EtherscanHashRate}, accessed \citedate{EtherscanHashRate})
@@ -670,7 +670,7 @@ A more realistic example for Proof Of Useful Work schemes is found in the field
 
 Finding these prime number chains becomes exponentially harder as the chain length is increased. That way, the difficulty in the Primecoin network is set and the difficulty is adjusted after each single block, targeting one block per minute \cite{Primecoin}.
 
-Primecoin is comparable to the GIMPS project (``The Great Internet Mersenne Prime Search'') which is run by enthusiasts on their computers all around the world and has found most of the largest prime numbers known to date.
+Primecoin is comparable to the \ac{GIMPS} project which is run by enthusiasts on their computers all around the world and has found most of the largest prime numbers known to date.
 
 %% \begin{frame}{Proof Of Useful Work: Renewable Energy}
 
@@ -748,7 +748,7 @@ Buterin is the co-creator and inventor of Ethereum, who has also written Ethereu
 The developers of Ethereum, Buterin amongst them, are currently planning a transition from pure Proof Of Work to a hybrid Proof Of Work-Proof Of Stake based blockchain. As of early 2018, this is still work in progress.
 \emph{Casper} is a smart contract that will implement and monitor Proof Of Stake on the Ethereum blockchain. \emph{Smart contracts} are small pieces of code run by every single participating node. Therefore, Casper (the Proof Of Stake scheme) is also executed and most importantly verified by every single node in the network.
 
-The first iteration, now called \emph{naive Proof Of Stake}, suffered from the \emph{nothing-at-stake problem}: it didn't punish participants for validating more than one history (i.e. more than one block), therefore creating splits in the blockchain and undermining its credibility. This problem has been addressed in \emph{modern Proof Of Stake} versions, such as ``Casper FFG'' (Friendly Finality Gadget, developed by Vitalik Buterin et al.) and ``Casper CBC'' (Correct By Construction, developed by Vlad Zamfir et al.).
+The first iteration, now called \emph{naive Proof Of Stake}, suffered from the \emph{nothing-at-stake problem}: it didn't punish participants for validating more than one history (i.e. more than one block), therefore creating splits in the blockchain and undermining its credibility. This problem has been addressed in \emph{modern Proof Of Stake} versions, such as \emph{Casper \ac{FFG}}, developed by Vitalik Buterin et al., and \emph{Casper \ac{CBC}}, developed by Vlad Zamfir et al.
 
 Both Casper versions use the same underlying concept: anyone can bond tokens (the generic term for coins), while decisions leading to a convergence on the blockchain are monetarily rewarded, decisions resulting in a split of the chain are punished.
 
@@ -785,7 +785,7 @@ A general concern with Proof Of Stake methods is often referred to as the \emph{
 %% \end{frame}
 
 \section{Delegated Proof Of Stake}
-This consensus mechanism was first implemented by the BitShares blockchain and has been updated several times since. Broadly speaking, it is comparable to the U.S. electoral college system. Under DPOS, the stakeholders can elect any number of \emph{witnesses} to generate blocks \cite{BitsharesDPoS}.
+This consensus mechanism was first implemented by the BitShares blockchain and has been updated several times since. Broadly speaking, it is comparable to the U.S. electoral college system. Under Delegated Proof Of Stake, the stakeholders can elect any number of \emph{witnesses} to generate blocks \cite{BitsharesDPoS}.
 Witnesses serve the role of validating signatures and timestamping transactions by including them in blocks, therefore they are BitShares equivalent of miners (without the intensive work, of course). Each account is allowed one vote per share per witness, this process is known as \emph{approval voting}. The top $N$ witnesses by total approval are selected, where $N$ is defined such that at least 50\% of voting stakeholders believe there is sufficient decentralization. When stakeholders express their desired number of witnesses, they must also vote for at least that many witnesses. A stakeholder cannot vote for more decentralization than witnesses for which they actually cast votes.
 The elected witnesses are paid for producing each block. The pay rate is set by the stakeholders via their elected delegates. If a witness fails to produce a block, then they are not paid, and may be voted out in the future, as they did not fulfill the promise to their voters. A single witness can't sign invalid blocks as the block needs confirmation by the other witnesses as well \cite{BitsharesDPoS}.
 
@@ -825,7 +825,7 @@ Advantages of Proof Of Burn include very low energy consumption, especially when
 One application of Proof Of Burn is bootstrapping a cryptocurrency. Users who ``burn'' the original coin (take Bitcoin as an example) by sending it to an irretrievable address are awarded with coins of new currency (i.e. Altcoins). As all transactions are stored on the blockchain, this process is fully transparent and can be verified at any time. This can be considered as a form of extreme, one-time investment.
 
 \section{Proof Of Luck}
-Proof Of Luck is the final consensus mechanism we take a look at in this paper. It was conceived by Mitar Milutinovic, Warren He, Howard Wu and Maxinder Kanwal from University of California in Berkeley. It utilizes the \emph{Trusted Execution Environment} (TEE / TXE) of modern CPUs as a resource, as there is only one per CPU. Accordingly, it relies entirely on the security and correctness of TXE provided by the manufacturer.
+Proof Of Luck is the final consensus mechanism we take a look at in this paper. It was conceived by Mitar Milutinovic, Warren He, Howard Wu and Maxinder Kanwal from University of California in Berkeley. It utilizes the \ac{TEE} of modern \ac{CPU}s as a resource, as there is only one per \ac{CPU}. Accordingly, it relies entirely on the security and correctness of \ac{TEE} provided by the manufacturer.
 
 It combines \emph{Proof Of Work}, \emph{Proof Of Time} and \emph{Proof Of Ownership}, and in their whitepaper the authors argue that it ``offers low-latency transaction validation, deterministic confirmation time, negligible energy consumption, and equitably distributed mining'' \cite{PoLWhitepaper}.
 
@@ -855,6 +855,29 @@ The applications for blockchains are vast and we might not have come up with the
 
 The current position of blockchain technology can be compared to the state of networked computers in the 70s and early 80s of the last century. Not a huge amount of value delivered at the time, yet very few could have foreseen the impact of the internet and personal computing would have on the world. If blockchains do present a new leap forward in how we communicate, the electricity used will not be wasted, as they are an invaluable tool for the future of distributed computing.
 
+\chapter*{\glossarytitlename}
+
+\begin{acronym}[ABCDEFGHIJK]
+  \acro{CPU}{Central Processing Unit}
+  \acro{GPU}{Graphics Processing Unit}
+  \acro{FPGA}{Field-Programmable Gate Array}
+  \acro{ASIC}{Application-specific Integrated Circuit}
+  \acro{SHA-1}{Secure Hash Algorithm 1}
+  \acro{SHA-256}{Secure Hash Algorithm 2, 256 Bits}
+  \acro{GIMPS}{The Great Internet Mersenne Prime Search}
+  \acro{FFG}{Friendly Finality Gadget}
+  \acro{CBC}{Correct By Construction}
+  \acro{TEE}{Trusted Execution Environment}
+\end{acronym}
+\clearpage
+
+\listoffigures
+\clearpage
+
+\listoftables
+\clearpage
+
+
 \bibliographystyle{IEEEtran}
 \bibliography{../resources/references.bib}