Highlevel concept comparison of Facebook Libra with Prasaga DataGrid Blockchain
Libra
Facebook just announced Libra and released both technical and business information regarding it. The technical side seems to consist of a leader-based, non-sharded, BFT consensus style, limited distributed ledger; coupled with a new language and VM with new built-in data types, and a “module” concept. The ledger is initially permissioned with a claim that it will become permissionless, while the ability to add new code is extremely limited, again with claims that it will be opened up in the future.
On the business side, there is the formation of a consortium of large companies that pay a nominal fee of $10 Million to become nodes running the Libra distributed ledger. All Libra coins are backed by some sort of basket of other currencies, and to obtain Libra coins, users are required to deposit some other currency (i.e. buy them), with one of these companies. The nodes operating the distributed ledger do not mint new coins (i.e. no incentive rewards). Coins are created on deposit, and destroyed on redemption of other currencies, presumably fiat (i.e. USD, EUR, etc.). This is known as a currency reserve concept. The reserves are held in deposits and presumably invested by the depository for their own financial gains.
The structure of the Libra Council (i.e. the consortium), appears to the following:
Contrast to Prasaga DataGrid Blockchaintm
The Prasaga Foundation structure is somewhat similar to the Libra Council, in that it operates as a non-profit trust. It has underneath it “The Institute” which is intended to incubate innovation for blockchain applications and uses. It also has the equivalent of a social impact board intended to act as a philanthropic investment body. Since the DataGrid Blockchaintm is permissionless and open, there is no role for an Association of node operators.
The Prasaga Foundation does not include a reserve. This is because the DataGrid Tokentm (“DGT”), instead of being a reserve currency, will use a monetary policy approach to manage the coin supply based on economic metrics of the accounts on the blockchain. Competitive market forces will drive the exchange rates between the DGT and other cryptocurrencies as well as fiat currencies. The DGB uses the incentives model for decentralized minting of DGT, which serves to manage the coin supply and help ensure distribution of the DGT.
Although the DGB concept requires an understanding of money supply, inflation and real economic growth along with an understanding of feedback control systems, it leads to an independent currency backed by the blockchain itself.
The DGB technology includes a new model for virtual machine code execution that is explicitly object-oriented, significantly increases the amount of code reuse, and supports language polyglot. The architecture of a DGB account enables sharding at the account level, as opposed to the smart contract level, greatly increasing the opportunity for parallel execution and resulting throughput.
Further, the blockchain account architecture offers opportunities for the creation of decentralized multi-tenant multi-sided markets, as an enablement platform, and means for implicit adherence to regulations such as KYC for investors, while providing an open, flexible programming environment for application developers.
In short, we believe the DGB offers open, permissionless blockchain technology, coupled with scalability, managed independent currency, advanced object-oriented language support, an enablement platform, and more.
Opinion
It seems to me that all reserve approaches I’ve seen so far basically enable a company like Facebook to take your fiat money hand you a token and then do whatever they want with your money. No FDIC insurance. No protection. A run on their bank would blow this up, leaving the token holders with unredeemable tokens. Although I’m sure the intent is to include safeguards against this, I believe the intent of cryptocurrencies was to avoid just this type of system.
I might also add they control the exchange rate to the basket backing their token. They can devalue it which means your token loses value at their whim. Again, I’m sure there are rules regarding the basket marked to, but it is still under centralized control.
The monetary policy of the DGB will not be under centralized control. Instead it will be algorithmically controlled with decentralized node incentives, with an oversite function that can modify the parameters of the algorithm. The oversite will be provided by stakeholder vote and thus be a decentralized open system. The greater the use of the DGB for economic activity, the more stable the DGT becomes.
Since the DGB account architecture directly supports sharding, the greater the number of accounts, the more opportunity for parallel operations, the greater the throughput capacity of the DGB. In other words, instead of a fixed number of transactions per second (TPS), the DGB can continue to scale as its usage increases. The Libra protocol appears to be a replicated database style with a distributed log. Although this has a high TPS, it is a fixed capacity independent of the number of accounts or transactions. Thus, I would expect it to have peak saturation issues and become unreliable over time.
Finally, Libra is using a version of Byzantine Fault Tolerance (“BFT”) with elected leader, which includes signatures of nodes and can be thought of as a Proof-of-Stake style protocol. Although there is significant evidence that the PoS consensus models relying on signatures using private/public key pairs, perform better on per-transaction basis, the reliance on the security of the private keys remains as a permanent issue. Further, without the compute intensiveness of proof-of-work, should an adversary acquire sufficient keys to over come the BFT at some point in the history, it is possible to create a fake chain and generate fake transactions at a rate faster than the “true” chain.
I do not believe that a blockchain can claim long term security without some aspect of proof-of-work included to eliminate the sole dependency on the private keys of the nodes. Therefore, the DataGrid Blockchain is designed to use both a form of proof-of-stake augmented with our form of proof-of-work. We believe that this gives our blockchain the long term security on par with Bitcoin or Ethereum, while provided scalability to support transaction throughput to current and future decentralized applications.
DataGrid Blockchaintm, DGBtm, DataGrid Tokentm, DGTtm, Extensible Blockchain Object Modeltm, XBOMtm, Extensible Smart Object Assettm, XSOAtm, Smart Objecttm, Extensible Signature Objecttm, XSIGtm are trademarks of Prasaga, LLC. All rights reserved. www.prasaga.com info@prasaga.com
