Aisioshechka from Zuckerberg - briefly and in the Libra case

    The Libra project announced on June 17, 2019 has already caused a stir in the press and even big politics, but all the articles I have seen are full of slogans and predictions of a grim future, omitting technical details. In this article, I will briefly outline what Facebook actually promises, how their tokens work and what the block frequency is.

    Economic model:


    • There is a certain Libra Investment Token, hereinafter referred to as LIT. Technical implementation is not indicated and has nothing to do with the matter.
    • You can buy this same LIT for a minimum contribution of $ 10,000,000. The issue is not limited, the charter of the association directly mentions the issue and sale of LIT in the future, after the launch of the project.
    • LIT holders are accrued all the profit remaining after the costs of maintaining and developing the project - essentially a dividend.
    • The unit of value is Libra. The price of Libra is declared tied to a basket of fiat currencies (through Forex), a basket has not yet been announced.
    • Libra is created in only one way - by depositing fiat money into the issuer's account, which then issues the corresponding amount of Libra.
    • The issuer invests the received fiat money in “short-term government bonds of states with a stable currency, a liquid government bond market and a high credit rating”. The interest on these bonds is officially the only source of profit for LIT holders.
    • Information on the current economic situation: the relevant interest rate for China is 3%, the USA 2%, Australia 1%, the UK 0.5%, Germany, Switzerland and Japan are negative.
    • The issuer itself also acts as a buyer of last resort, in fact undertaking to exchange Libra back for fiat money on demand.

    From the above, we see that the more Libra is released, the greater the potential profit for LIT holders.

    Organizational Model:


    • Libra Association, registered in Geneva and with employees in San Francisco. Not directly related to Facebook. The charter of this company is described in some detail in the project documentation. He is involved in making decisions regarding the blockchain, paying dividends and public relations.
    • LIT holders have the right to vote, but it is limited by a number of exotic rules (maximum 1% per representative, maximum 33% from non-profit organizations, etc.).
    • Calibra is a Facebook subsidiary implementing the first wallet and other technical issues - including KYC and legal aspects in different countries.

    Most of the charter is written in a strange language confusing physical people voting in the Association, validator nodes voting for blocks, token holders and some Founding Members. There are points openly in conflict with the protection of minority shareholders.

    Technical Details:


    • Surprisingly technically conservative product, a kind of blockchain from 2015 without frills. There is already a testnet, the client is free, do not want to start it. For people familiar with the air in the project is not unusual.
    • Blockchain based on PoA, everything is simple and without risk. From a well-known and closed group of validators, a leader is appointed who offers the block, validators vote for the block by signing with the keys supplied off-chain. The same validators are sync nodes, archive nodes, and generally the only full members of the network. Hello from 2015.
    • The charter describes simple and understandable criteria by which validators will be selected among those who wish. The minimum technical characteristics of the necessary iron are indicated: m5.24xlarge or half-rack if hosted independently.
    • Initial number of validators: 100. Transaction finalization time: 10 seconds. Estimated throughput of 1000 tps.
    • Yes, over time, I certainly deceived you. Blocks in the system are present only in the context of the consensus protocol and are not visible at the data model level. The role of “block depth” here is played by the “database version”, which is nothing more than the number of finalized transactions. In practice, I do not see a difference, such a model is also quite logical.
    • There is a concept of gas that is paid at Libra. The semantics are the same as on the air: gas price and gas limit, i.e. Turing-complete smart contracts. Unlike the ether, there is no separate concept of “just transfer” - Libra translation is also an invocation of the corresponding smart contract.
    • Briefly about smart contracts: they promise a new language Move, but for now you need to use the language Intermediate Representation (IR), which is very similar to (shock! Sensation!) Solidity. Users cannot deploy smart contracts to the state database; only persons designated by developers can do this. A transaction is essentially a smart contract consisting of one function that launches the public functions of already deployed smart contracts. On the one hand, of course, a closed ecosystem, on the other hand, it is simple and elegant, and to some extent completely logical.
    • There is a simple model by which non-validators communicate with the network. Everything works on the basis of queries - we send a request in a special language and we get a validator response. That is, no synchronization, no light clients, question and answer.

    Other important details:


    • Wallets and exchangers will be under KYC, but the blockchain itself is open and the client is available. KYC information is not in the blockchain itself, actions with which (judging by the testnet) can be carried out directly. Do you see the problem? I'll tell you: “Oh, my phone was hacked” and you get an anonymous Libra.
    • The charter directly states that closing fiat transfers at the issuer level is a valid way to select a fork. I appreciate the directness and uniqueness.
    • All the documentation is riddled with promises of transition to PoS, where Libra itself will be the guarantee, but the plans are rather vague: “start the transition process five years after launch”.
    • The technical documentation is written professionally, but in very simple terms - starting literally with what keys are, what transactions are, etc., it is obviously designed more for schoolchildren or students than as a guide to bearded uncles integrating another token into the exchanger.

    A source

    Also popular now: