 reluctantly crouched at the starting line, engines pumping and thumping. You're watching The Daily Decrypt, a place where there are only two kinds of people, those who like currency competition and those who really like currency competition. I'm your host, Amanda B Johnson, and today's episode is brought to you by the BTC subreddit. As Bitcoin Core Development has moved further away from the parameters outlined in the original white paper, namely that of an electronic peer-to-peer cash system, many have wondered if Satoshi's vision is even feasible anymore at all. A post was recently made on the Dash subreddit entitled, Dash Is The One True Bitcoin as Satoshi Envisioned It. The post's author claims that Dash's tech specs fulfill Satoshi's dream of electronic cash, and in addition offers the governance to make that possible. Is this claim credible? Well, you can judge that for yourself, as I'm about to put on my geek glasses and read for you the Dash Evolution white paper. All right, sit back and relax if you like to just listen along, or if you'd like to read along as well, I will place the text on the screen. Let us begin. Dash, version 13, also called Evolution. Design Overview, revision one by Evan Duffield. Abstract. Dash, an open-source cryptographic currency derived from the Bitcoin protocol, added a second tier to Bitcoin's P2P network architecture, masternodes, that enabled Dash to improve on the decentralization features of Bitcoin with new capabilities such as trustless transaction mixing within the second tier network for privacy, dark send, instant transaction confirmation without a centralized authority, instant X, and decentralized governance and funding by blockchain, DGBB. Other features added include multi-phased forks, sporks, a chained hashing algorithm, X11, and a mining difficulty adjustment algorithm to address flaws in Komodo's gravity well. Dark gravity wave. In the third major iteration of Dash, named Dash Evolution or version 13, additional architectural and functional improvements are being developed such as the addition of a third network tier, T3, comprised of a decentralized API, DAPI, that provides users with trustless remote access via direct HTTP and RPC connections into the Dash network that are serviced by randomly comprised masternode quorums. A decentralized wallet protocol that enables users to buy merchandise from the web in a trustless way, Dash Pay, without the need to host their own full node or use a centralized payment gateway, a second tier high-performance shard-based file storage system that provides improved methods for transaction confirmation and double-spend prevention, Dash Drive, primitives for representing users and accounts as objects to enable users to connect and transact with friends using aliases and rate each other to build trust networks, social wallet, decentralized network administration by masternode operators, DNA, a new dynamic query language for use across Dash Evolution components, providing an extensible object-based cross-tier communications standard, DSQL, addition of a historical chain of all signatures used on the masternode network for use in secure quorum selection, quorum chains, improved privacy architecture, automatic instant transactions, Auto IX, and rated services provided by network operators such as Fiat Converters. Oh, just the abstract. Let us dig in. Number one, introduction, 1.1, birth of Dash, a new P2P network architecture. Dash, which stands for Digital Cash, was originally known as Darkcoin and launched in 2014 as a privacy-centric cryptocurrency derived from Bitcoin. Dash implemented the first decentralized privacy system, DarkSend, onto the Bitcoin protocol based on CoinJoin, the most widely used anonymization technology for Bitcoin. CoinJoin works by joining different users' transactions together in set denominations to obfuscate the transaction history. The main problem with CoinJoin is that users need to negotiate transactions they wish to join, and there's no known way to do that within the Bitcoin protocol, meaning users have to send their funds outside of the Bitcoin network to centralized CoinJoin services and lose control of their funds during the process. Dash wanted to create a decentralized implementation of CoinJoin, where users can anonymize their transactions without losing control of their funds to a centralized intermediary, and the solution concept was to expand the Bitcoin P2P network architecture to enable full nodes to provide the transaction mixing service internally between randomly selected groups of full nodes working together in series. An issue with this concept was that full nodes are relatively inexpensive to run. This means the system would be susceptible to Sibyl attack, because malicious parties could create thousands of full nodes at low cost and game the system by controlling the large percent of full nodes required that could then act maliciously to gain centralized control of users' transactions during mixing, and potentially record enough transaction joins to de-anonymize a user's transaction history. This led to the solution of requiring users to prove they had a certain level of collateral to be able to provide this full node mixing service, meaning gaining a majority of full nodes would become prohibitively expensive. Dash names these collateralized full nodes, masternodes. The second issue with the design was that it is more expensive to run a full node that is providing a mixing service, so there is no incentive for users to do so, other than for altruistic reasons. But to be reliable, network services need to be highly available, with enough nodes to provide an effective service, so Dash's solution was to extend Bitcoin's principle of rewarding miners with blocks of newly created funds autonomously on the blockchain, to also reward masternodes that provided the mixing service honestly. This enabled Dash to create the first reliable and decentralized way to anonymize transactions in any Bitcoin-based cryptocurrency, and at the same time incentivize the infrastructure that provides the service. This has been highly successful for Dash, with no dark-send transactions ever shown to have been de-anonymized, positive code review and adoption gained in privacy-focused markets along with Bitcoin, and with Dash's second-tier network growing now to over 3,300 masternodes, meaning Dash is second only to Bitcoin in terms of transaction throughput capability. 1.2 Phase 2 New Capabilities With dark-send in place, the Dash creators realized that the second-network tier of masternodes could provide a lot of additional benefits over and above the original Bitcoin network architecture and new decentralized functions and services that had not been possible before could now be created. Now that Dash had a high-performance, high-availability second-tier on the network that was growing rapidly as users upgraded and collateralized their full nodes and masternodes to get paid for network services by the blockchain in the same way as miners, the creators started innovating new technologies using the new capabilities of the network. The first innovation on top of the second-tier network architecture was a method to provide instant transaction confirmations, less than 4 seconds on average, through transaction locking and masternode consensus called InstantX, enabling users to pay instantly with Dash. The second major change was to reallocate a percentage of the mining block reward, 10%, and allowing masternode operators to determine to whom it's paid for jobs that people propose to the network, with secure decentralized voting on the second tier by using the collateral public key from each masternode to sign messages from each masternode, thus creating a decentralized funding mechanism which is permanent and can care for the ecosystem for the very long term. Dash names the system Decentralized Governance by Blockchain, DGBB. And since DGBB's release in August 2015, all Dash operations have been under the control of the Dash network via the DGBB system, with new proposals made and funded each month autonomously by the network. In fact, since the inception of this system, the Dash network has funded the promotion of Dash conferences around the world, acquired high-value property directly from the blockchain, Dash.org, and many other projects that were important to the long-term success of the ecosystem. We believe that Bitcoin is in fact the first ever decentralized job network to exist if we define jobs on the Bitcoin network as actions by network participants who are paid directly and autonomously from a purely decentralized source such as the block reward. Dash unlocked this concept by adding a second job, the masternode operator, which runs infrastructure for the network and is compensated. These masternode operators exist in a secondary market within the larger primary mining market and have an equilibrium with the amount of capital available to fund the infrastructure. With a new type of scalable P2P infrastructure and these new decentralized technologies in place, we have identified several new changes to the network architecture that we can make to provide a new level of capabilities to bring cryptocurrencies closer to mainstream users, and we are naming this iteration Dash Evolution. The first major change in Dash Evolution is a decentralized API, DAPI, that is implemented by using small processing groups of masternodes hardened by proof-of-work randomization. Users can execute simple requests over HTTP, which update or retrieve their information, and these small clusters perform the work as a group action, executing the event many times and signing the result. This opens up a cryptocurrency, for the first time, to decentralized and trustless use directly from the worldwide web, without reliance on any individual intermediary to handle users' funds. Merchants can integrate Dash Revolution into their e-commerce store as easily as they do with centralized services like PayPal or Google Wallet, whilst receiving a faster and cheaper service that lets them transact directly with their customers without having to use an intermediary payment gateway or face increased cost and complexity to integrate and maintain their own full node infrastructure. Users can check DAPI result signatures against a deterministic quorum chaining algorithm similar to that of the blockchain, but designed to preserve the historical and current state of the masternode quorum groups on the network. The quorum chain begins with one masternode, the source masternode, signing a message, which adds the new masternodes to the next quorum chain entry, and so on. As long as the user has the source masternode key, they can independently check the masternode quorum chain in a secure and efficient way. By requiring each second-tier node to have access to a specific amount of storage on the network, we can begin to create a decentralized storage system capable of vast storage capacity. This can allow the storage of metadata and other network-related data to be stored on the network on behalf of the users, which can then be retrieved privately at any time in the future. Rights to this file system are executed in a similar way, requiring all or most of a quorum to agree on what is being changed in the decentralized file system. Using DAPI and the decentralized file system, we create easy-to-use primitives such as users, groups, accounts, and features such as private messaging, group messaging, and name-based payments. Users can move from device to device and by querying the data stored privately on the network, automatically syncing all of this data, allowing for a streamless, centralized seeming design, which is in fact decentralized start to finish. Evolution also consists of a lot of other new features to make cryptocurrencies easier to use and access to mainstream users. I'm sorry, let me rephrase that. Evolution also consists of a lot of new other features to make cryptocurrencies easy to use and access to mainstream users, I was just poorly worded, and whilst keeping Dash fully decentralized and scalable, which are given an overview in this document and more specific details and separate documents for each major area. 2. Multi-Tier P2P Network Architecture In most P2P systems, all nodes on the network use the same software and have the same role. We expand on this concept by introducing differentiated roles for the nodes on Dash's P2P network. The different roles allow the network to be increasingly robust and self-sustaining. Nodes can do many things on the Dash network such as running a collateralized infrastructure and voting on proposals for governance and funding. These proposals in turn determine the direction of the currency and the network in a decentralized way. Miners form the first tier of the network. Their job is to create and validate a permanent and immutable record of transactions by creating blocks in Dash's public blockchain. For this, they are paid a portion of the block reward. Miners also ensure that approved budget proposals and masternodes are paid. These masternodes form the second tier of the Dash network and provide vital infrastructure, including storage and processing power for the network. The third tier runs on external hardware, such as browsers and phones, and connects to the Dash network via a decentralized API, DAPI. 2.1, first tier, mining and full nodes. The first tier of the Dash network consists of miners and full nodes. Miners act normally, as on any proof-of-work-based P2P network. In addition, we have the ability to support undifferentiated full nodes, which can validate transactions and keep up to date with the current state of the blockchain. Miners share reward with the rest of the miners according to their contribution to the network itself. In addition, they must honor additional payments out of the blocks they work on via connecting to the daemon and checking the block template. These extra payments service the network by paying for the second tier infrastructure and decentralized funding mechanisms. 2.2, second tier, infrastructure. A stable infrastructure is provided for the network by special nodes we refer to as masternodes. These are full nodes on the network that are required to meet specific minimum requirements with respect to hardware, configuration, such as certain ports open, etc., and participate in regular software updates. By utilizing collateral transactions, unspent input signed by the operator for actions, we can organize a secure group of users to provide services for the network, allocate funding, and govern the currency. The second tier is essentially its own market within the larger mining market. As a market-based solution, a natural equilibrium is maintained between the number of masternodes running and the payment each masternode receives from the blockchain. To provide services for the network, the second tier operators run specific software, the normal dash daemon configured specifically for masternode operation, which allows the users to connect to them using a decentralized API, DAPI. When connected to this DAPI, users can do various things on the network like friend others, update their profile, send money, and so forth. These features use the new dash drive system, which allows secure reading and writing to the decentralized file system. 2.3 Third Tier Users The third tier is comprised of the end users of Dash. They have no direct access to internal communication happening on the network. Instead, they connect to the network through a DAPI, which allows them to receive secure, fast, and lightweight access to the network. By using a REST API, we can allow low resource, friendly access to the network. To access the features of the network, users will connect to a round robin DNS, which points to the entire second tier masternode network. A simple open source script that takes the masternode list and updates DNS entries will be available to anyone who wants to provide this service. T3, Tier 3, developers will then be able to use many known DNS entries in their software for robust, decentralized access to the network. After connecting to the network, users can submit tasks they would like to execute using a simple REST-based interface. As an example, accepting friendship. Random groups of masternodes are selected by the network to simultaneously perform the work users ask to be done. These groups are known as quorums. After doing the work, each member of the quorum will sign and return the results and signatures to the end user. Users can then be sure the work was performed correctly without tampering. 3. Decentralized API Users will be able to buy goods directly from websites without using any centralized intermediary services. We seek to serve many users in a completely decentralized way using only the second tier, operating similar to a back end as a service system, BAAS, but fully decentralized. Merchants will use an SDK when accessing the Dash network from their site. This SDK enables users to click a buy with Dash button, which opens a connection to the masternode network and makes a payment via DAPI. There's a graph of the process. Should you care to look at it? And here is another graph, 3.2 of the DAPI internal resolution. Should you care to look at that? 4. Dash Pay. It says please refer to the document Dash Pay Decentralized Wallet. We don't have that here, so we'll continue. 5. Social Wallet Users will have a username and password, which they use to access the network. The username will be reserved by using features of Dash Drive, and the password, 8 to 12 words, will be used to create an HD Wallet. A primary key will be pulled from this HD Wallet, which is used to retrieve information about the user from the network. All communication with the network is protected by these root communication keys. To assist in adoption, users will utilize wallets with constructions designed to make working with the network much easier. The goal is to provide an experience similar to a centralized website, but without any centralization. Users will maintain a friends list on the network, exchange future keys to use privately, create groups for companies and other uses, and create shared accounts so that users and funds can be added and removed. With this combination of features, we can achieve a level of user experience that was previously reserved for large-scale centralized systems, with the added benefit of decentralization where users keep control of their funds and don't need to trust any central entity or be exposed to any single point of failure. Users will also be able to rate others on the network in a nearly permanent way. Once two users have done business with each other, they may add feedback to each other's profiles. This helps future users make informed decisions based on reputation. Other notable changes include the ability to send immutable transactions instantly and privately using advanced features of Dash Drive. All transactions are instantly confirmed so that users don't have to wait long periods for multiple confirmations on the blockchain. All transactions that are successfully written to Dash Drive will be successfully archived in the blockchain. 5.1 Fiat Conversion. Dash Evolution will also give users the ability to convert to and from fiat using converters. These are special users that provide another service for the network and are paid a percentage of the transaction to process it. Converters will be rated by users who do business with them and this reputation will be crucial in helping other users decide which converters are most reliable and trustworthy. The results will be a safe and global fiat conversion network. 6. Dash Drive. Each member of the second tier will be required to have a specific amount of storage space in order to power the Dash Drive file system. By sharding the storage via the collateral transaction hash, we can define 1024 different shared storage devices on the network. We use 1024 because we can identify shards by using the first 10 bits of a unique hash per storage object. For example, with a 40 gigabyte allocation requirement, the network can enjoy about 40,960 gigabytes of storage space. When users interact with the network, they will transmit information to be stored on Dash Drive via the DAPI. For redundancy, each shard will be stored multiple times on the network. For example, if the network has 5,000 masternodes, we will store each item 5,000 divided by 1024 plus seed count times. Dash Drive supports a few advanced features such as transactional commits where users can require multiple files get written to different destinations on the network. If any write fails, the entire commit for all files will be reverted. In addition, reading or writing files is only possible when a user has access to a given file, such as their own profile page. When trying to read files a user does not have access to, they will be denied access. Writing files can be done only by having enough quorum signatures and can be used to do maintenance or allow users to update information on the network. 6.1 Masternode Objects Masternode objects are stored on all shards on the network for quick access from any node. There are only a few thousand of these servers maxing out at 21,000, so the file overhead will be minimal. Masternode keys will constantly be used on the network, so these files will be cached by the software in RAM, creating a masternode list functionality. 6.2 Quorum Chains Quorum chain is a historical chain of all signatures that have ever been used on the masternode network. This historical record allows us to check all signatures that have ever been used on the masternode network and allows the network to decide which masternodes are qualified to be in a quorum at what time. Calculation of the quorum chain uses a random number generator seeded with a historical proof-of-work hash to sort an array of masternodes. Masternodes are simply added and removed from this list by writing additional files to dash drive by majority quorum. 6.3 Users, Groups, and Accounts All dash drive objects are binary files which are written to dash drive on one of the 1024 shards. The files are written as a vector of integers and char vectors which store fields for user data. This allows users to have many fields such as username, password, friends, etc. 6.4 Doublespend Prevention, Commit or Roll Back Doublespending is not possible on the dash network due to the Commit or Roll Back C-O-R feature of dash drive. After sending a transaction to the network it will be written to dash drive and the inputs will be reversed via usage of the file system. If at any point a right to dash drive fails the earlier rights will be reverted back to the state before the commit started. This allows two users or one attacker can attempt to write a transaction inputs being that inputs are unique file locators on the network. Only one will successfully be able to write the transaction commit. In addition to reserving resources on the network dash drive stores a whole transaction history across the shared file system while awaiting archival in the permanent blockchain. Here's an example. C-O-R Operation Commit or Roll Back 6.5 Guaranteed Resources, P-O-S-E. This section is currently being written or rewritten. All right hop on. 7. Decentralized Network Administration The dash network is run by a decentralized administration workforce known as the Masternode Network. This is a decentralized P2P network comprised of thousands of people that together can administer, remove threats, deal with attacks, and reduce the load and strain on the network by taking appropriate actions. In version 13 the developers have a whole new toolkit for interacting with the currency as well. All users on the network will act under a single private identifier or username. This username is able to carry historical transaction ratings for each transaction it completed. This can be thought of as an eBay-like feature where merchants carry historical ratings and users base their business dealings on those ratings. Eventually ratings will be able to analyze using a graph similar to PageRank. By utilizing an algorithm like this we can build the first trust network on top of a decentralized currency. The network can automatically calculate these trust numbers via Masternode Quorum maintenance and assign them to user profiles. Users will then do business on the network according to these numbers. Based on the shared infrastructure in quorums the Masternode operators will use a SQL-like language called DSQL. This will allow the Masternode network to vote, to ban malicious users, or even entire networks of related attackers. After ridding the network of these attacks we can restore the trust network to its pre-attack state. In the example of a DashDrive data write attack the Masternode operators will be able to rewind the DashDrive changes on the network to reverse the damage. They can also prune and undo data adding attacks on the blockchain. This would be done using majority quorum actions where a large percentage of the network was required to vote making it a highly secure operation. Transactions on the network are free and users will have a certain limited amount of network processing time they can use. Larger users will be able to utilize additional processing time on the network by paying a fee that is added to their account. By definition network administrators run Masternodes for the network on the second tier. Each year they prove they own one Masternode which is recorded in their file. They can then vote on executing various commands on the network to administer it. For example, there might be votes to ban malicious users or nodes from the network. 8. Privacy and fungibility. Note the full details of this new system will be discussed in a separate paper. 9. Automatic Instant Transactions. Since the introduction of Instant X, Dash has enabled rather has established itself as the fastest cryptocurrency on the market. However, in the current system version 12, this service has to be requested in the wallet. This will no longer be necessary as the latest implementation of Dash will provide an automatic Instant X service for all transactions on the network. This will also be covered in greater detail in the Dash Drive paper. 10. Decentralized funding and governance. Decentralized funding and governance of the currency are very important foundational components of a system that is sustainable and can scale without issue. To achieve this we must add a form of decentralized funding for paying for things like the website maintenance, core development, business development and promotion. Decentralized financing is implemented by adding a proposal and voting mechanism to the Masternode system. By adding such a system we have a democratic, smart and efficient funding system that will leverage our Masternode operators to pick which budgets to fund and which not to fund. As Masternodes change hands the control of the currency will move with them so there will always be a decentralized group of people watching over the currency and network. As a network we can choose to finance anything we want such as large advertising campaigns targeted to specific regions or demographics for attracting users or new developments. We can develop specific markets in a specific order to have the greatest effect. Good decentralized governance of the currency is very important since it gives the user base and investors a clear sense of direction. In the case of Bitcoin there is no clear system of governance and as a result chasms have formed in the core team and in the community. To avoid that we give each Masternode operator one single vote on each proposal. We can then use this system and take multiple votes about specific options for advancing and improving the currency. This is all done in a completely transparent and trustless way. The decentralized budget system will be well funded because the Dash protocol ensures that 10% of the total currency emission every month is automatically redirected to support the ecosystem without relying on centralized sources of financing or altruism from the community. Funding is very important because as a project gets more important it will require better financing to remain competitive and to be properly maintained. Here's a budget funding per year based on potential market caps. As you can see our financing for the project grows as the value of the Dash currency increases. This results in a virtuous cycle. Decentralized funding goes to finance network development which creates price appreciation. This increase in the value of the Dash currency makes more funding dollars available to further enhance and develop the network which drives more currency appreciation. This will allow us to completely finance development marketing legal and infrastructure costs throughout the life of the currency. The budget system is also real time so Masternodes can trim the fat to keep the budget as lean as it can be 24-7 and votes can be changed at any time. In the event that a project with a recurring budget fails to deliver the appropriate results Masternode owners can always change their votes to no and if enough Masternode owners do so that particular project will stop receiving funding. 11. Masternode Quarums. Oh there's a cloud. When we talk about information in the first tier of Bitcoin Core we talk about processes that all clients on the network can check and verify. These must be locally verifiable things so that all nodes can do the same verification. If one node rejects something all nodes should also take the same action when checking the data. This is the basic idea of how Bitcoin operates however this type of logic is quite limited. Nodes can only provide information that they directly have their hands on such as is this signature correct? Is this hash correct? Is this proof of work correct? Does this script check out? If these things are true in one place in the network they are true in all places. This type of logic is obviously limited. We propose using a technology called Masternode Quarums which will be used extensively through Dash in order to provide a robust system that has never been possible with other types of logic. Masternode Quarums are powerful because groups of Masternodes can now do tasks that might not be possible to execute over the entire network at once. If the 10 Masternodes selected for a Quorum are successful at getting the same information then the entire network can consider that information true after that point. To stop Masternodes from colluding we base the deterministic selection algorithm on the proof of work hash. This provides an unforgeable base for a secure set of Masternodes to provide Quarums. As an example, Alice issues a command to get the current price of the stock AAPL for an on blockchain bet. She bets that on December 12th, 2018 AAPL will be greater than 3,000 US dollars. When this time passes the network takes a past block hash then using deterministic math the whole network selects the same set of Masternodes for this task. 10 Masternodes then become active and hit a public API on the internet. When these Masternodes are activated they trigger a Python script which has a list of APIs. We can do things like query stock prices, check the news, temperature, RSS feeds, etc. Broadcast transactions on the Bitcoin network, Litecoin network, etc. Download a web page and search for a piece of data in it. With this you can ask things like on google news page dot html does it have news item one? The answer will then come from the network. Another example usage is InstantX where we ask 10 Masternodes is this transaction valid and can you lock the inputs on the network so double spending is impossible? The network then either answers yes or no. This also allows for the Masternode network to serve as a decentralized oracle about anything and add scripts to transactions that reference their answer and contain their signatures. 11.1 Masternode Quorum security. The probability of winning the election will be one in n Masternodes. Currently the network is supported by over 3300 Masternodes. Each Masternode has a probability of one in n of winning the election. Therefore to attack the network the election process must select all of the attackers Masternodes. We will consider an attack on the transaction locking system by purchasing Masternodes in order to rig the voting system. For simplicity we will use a network consisting of 1000 Masternodes. Probabilities of attack can be calculated by the chance of a Masternode being selected as the winning node for a given block, one out of 1000. To subvert the system an attacker would have to operate all 10 Masternodes that won a given election. At a cost of 1000 dash per Masternode it's prohibitively expensive to attempt to attack the transaction locking system. To gain a probability of 1.72% of being selected for a specific block one has to control two thirds of the Masternode network. To gain control of two thirds of the network an attacker would need to purchase 2000 Masternodes requiring the purchase of 2 million dash. In this graph where n is the length of the chain of Masternodes, t is the total number of Masternodes in the network, r is the number of rogue Masternodes controlled by the attacker and is greater than or equal to n and the selection of Masternodes is random. Considering the limited supply of dash 6.03 million at the time of writing and the low liquidity of its market it is virtually impossible to obtain a large enough supply to succeed at such an attack. In the case of an attacker attempting to rig the voting system in favor of the wrong transaction lock i.e. the lock that isn't propagated across the rest of the network, the network will form an irreversible lock causing the transaction to the merchant to be invalidated. The merchant's client in question will permanently show an unconfirmed transaction due to a double spend and will never show the transaction was instantly validated. 11.2. Tier 3 Masternode Quorum Actions section currently being written. 11.3. Decentralized Governance of Block Reward Allocations. On the current dash network there are defined allocations of how the dash block reward is split between payments to Masternode operators, miners, and decentralized budget system. These were defined ahead of time and might not always meet the needs of the currency in the future. Masternodes will be able to form majority quorums using specialized commands to vote on changing these numbers when required on the network. Decisions to change key parameters will require an overwhelming majority such as 70 to 90 percent of the votes actively voting on such a change. This will allow the Masternode network to add more Masternode servers if the load on the network is too high or add more decentralized funding if the network requires that. 12. Scalability and Performance. 12.1. Incentivized Scaling. Within the dash ecosystem we have an incentivized Masternode network that carries the load of the network. These operators are properly incentivized so we do not have to worry about block size limitation or the bloat that would be caused by allowing millions of transactions a day. A larger blockchain simply means that Masternodes will need to be hosted on servers with greater storage capacity. The increased cost of renting these servers will be more than offset by the price appreciation that comes from mass adoption. When limiting the number of Masternodes by requiring specific amounts of collateral to launch one, the network is therefore limited to an easily calculable maximum number of Masternodes. For example, with 5.5 million dash in existence and 1000 dash in collateral required per node, we have an absolute maximum Masternode count of 5500. If we expect that 50 percent of the coins in existence will be used to collateralize Masternodes, this would give us a soft limit of about 2750 Masternodes, 2750. If each action requires 5 Masternodes to do the work in parallel and each action takes 10 milliseconds, we can determine the amount of work the network as a whole can do before being overloaded. Using these assumptions, we can demonstrate how the Masternode network will allow us to scale indefinitely. When starting a client that uses the decentralized API, the client will complete a series of 10 to 25 network actions depending on what services the client is using. It follows that if the network supports 275,000 actions per second, APS, that should allow between 15 million and 45 million users to utilize the network's services. As the network scales up, we can achieve an upper cap of nearly 1 billion simultaneous users. 12.2. Access speed over time. With all user data stored on T3 and accessed via secure API, the accessibility of the service will never begin to suffer from too many users due to blockchain bloat. A larger blockchain does not mean longer sync times for any devices connected to T3. Accessibility to this system will largely be determined by the state of the hardware in T3 and the number of users currently hitting the API. 12.3. Fee structure, block size, spork protection. We believe that an incentivized full node infrastructure, decentralized governance and funding will allow us to have no block size limitation for unlimited growth. We don't want to hinder this growth by external limitations that were designed to protect the network from the incomplete incentive models used in the Bitcoin ecosystem. Fees will be kept near 1 cent per transaction or less, regardless of the value being transferred. This opens us up to the possibility of spam attacks. So as a precaution, the network will be able to trigger a multi-phase fork, spork, to stop spam attacks on the fly. The network will do this by automatically raising the minimum fee for transactions until the attacker runs out of money. After the attack has been deflected, the network will return to normal operations. Such a solution is only one option. Masternode operators could eventually reverse this type of attack by using majority quorum commands to clean up the blockchain by deleting the spam then ban the users who performed the attack. 13. Improvements to decentralization. When designing a cryptocurrency, we should strive to have as little centralization as possible. There are many places a cryptocurrency can become centralized and thus prone to being attacked. The most important areas that we've identified are one, a centralized database, ledger. This is the problem Bitcoin solved by decentralizing the ledger in a trustless way. Two, core development, dev. Must be open source and not centralized to the will of a few powerful individuals, i.e. decentralized governance via voting to make decisions. Number three, infrastructure. Infrastructure one. If the network hardware is mainly provided by one company or person that leaves the network prone to monitoring and weak to attacks. Number four, lack of infrastructure. Infrastructure two. If the level of hardware providing public service is not great enough to support the users, that can leave the currency prone to DOS attack. Make sinking slow and create other similar problems. Number five, lack of decentralized funding. Funding. If a project is receiving its financing only from centralized sources, that can that can leave a project weak to coercion or undue influence from these sources. Six, centralized mining pools. Mining. One issue that's always plagued cryptographic currencies is the rise of pooled mining, which leaves the system weak to attack by a party or small group of parties that has more than 51% of the hash power. Number seven, decentralized decision making decisions. By not having a clear way to resolve disputes, the currency is subject to stagnated development or control of a minority party. Number eight, centralized money transfer, empty. If all points of entry into the ecosystem from fiat are controlled by centralized companies, that's a weakness as well. It is clear that when developing a cryptocurrency, there are many weak areas that must be eliminated in order to make the currency more robust and less prone to attack. We want to design a currency that isn't weak to any of these types of attacks. You see here a graph of the points of weakness in various cryptocurrencies when compared to one another. Dash is not perfect in all of these areas, but we're trending toward a much higher state of decentralization. In the rest of the documentation, we touch on improving these weak areas in the coming years to design something that's almost completely decentralized. Fourteen, design improvements. There are some known issues with this design that can possibly be addressed before the launch of Dash Evolution. This section will serve as a starting point for brainstorming additional improvements to the technology. When sending money to known aliases on the network, the payments are completely public and transparent. At all other times on the network, addresses are provided to all parties in a transaction via encrypted private messaging. Even public transactions are anonymized in the second phase of blockchain archival, so the transactional mapping would need to be stored permanently and sniffed off the network at the time the transaction was sent and off the specific shards involved. Fifteen, conclusion. Bitcoin is a revolutionary currency and a brand new way of thinking about money, but we believe there are parts we can improve on. We introduce a series of design changes to the foundational code, which attempts to correct the incentive issues with the Bitcoin network and thus helping scalability, governance capability, funding mechanisms, and ease of use. Dash solves these issues by a, allowing scalability by incentivizing the network in various ways, b, decentralized funding to finance operational requirements of a currency without risking the decentralized sovereignty of it, c, decentralizing governance resulting in a clear, decentralized, and trustless decision-making process, and d, easier, internet-friendly access via the first DAPI for anyone to access the network without having to run and sync a full node or rely on centralized services. Since all users will be utilizing the services through the third tier, we simply need to make the interface they use static, and our network can change radically on the inside without disturbing worldwide commerce outside of the Dash network. This allows us to try adventurous new ideas and update the network frequently without disrupting the user experience to keep Dash as a leading innovator in decentralized cryptocurrency technology. The end. Today's episode is brought to you by the BTC subreddit, a community that isn't afraid to talk truth even when that truth might hurt. Our BTC's traffic has shown considerable growth in the last few months, attracting those who want to talk Bitcoin without the censorship. You can join the conversation yourself at reddit.com slash r slash BTC. And the daily decrypt studio is soon to get a makeover. Now, much as I know, you adore carpet padding. Get your fill of it now while you can, because changes are coming. We've reached out to lead developer Evan Duffield. So yeah, it uses these decentralized forms. They could also be thought of as decentralized oracles, because they can pretty much do anything, but they have to do it together.