Thread by @treeder

The problems with blockchains, how to fix them and announcing a new project. A long thread…

🧵

Blockchains have obvious problems such as cost, energy consumption, and scalability. There are other issues too such as usability, safety and security of personal funds. Let’s discuss all of these:

Cost:

Most blockchains are far too expensive to use for day-to-day spending. Ethereum fees right now are roughly $5 to send ETH, $12 to send USDT, $45 to do a Uniswap swap, and $50 to buy an NFT on Opensea. These prices change by the minute.

None of that works for any normal person trying to use crypto for payments. If you have to pay twice as much for your @netflix subscription, that clearly isn’t going to work. The only plausible reason those are acceptable costs is that you think you’re going to make more...

... money than you lose to those fees. Which is the reason most people have crypto… currently.

And worse, those fees aren’t creating any value. They aren’t going to Netflix to make more and better content, they are going to miners to make them richer. This creates a perverse cycle of incentivizing more mining and therefore using more energy and so on.

If you’d like to learn more about how silly and wasteful the proof of work process is, I’d be happy to explain.

Energy consumption:

Bitcoin and Ethereum energy consumption is frankly ridiculous. Together, they use roughly 1% of the world’s entire electricity supply. Read more: medium.com/gochain/the-tr…

Proof of Work will not work in any sustainable future scenario. Many new Blockchains use Proof of Stake (Ethereum has been trying to switch to it for years now). Proof of Stake is much more energy efficient, but has its own fair share of issues.

Scalability:

This is where the real problem lies because it affects almost everything else including cost and usability.

On the slow end of the spectrum, there is Bitcoin and Ethereum which can handle about 7 and 13 transactions per second respectively. That’s for everyone in the world to share and it's barely enough for one app, let alone ALL the apps.

Some newer blockchains can handle orders of magnitude more tps, but still nowhere close to enough. And they ALL have fundamental technical limitations, like database scaling 101 kinds of issues.

The data structure of a blockchain is… well… a chain of blocks. Within each block is a chain of transactions. Which means you take a bunch of transactions and wrap them up then add that to the chain. All the transactions must be in sequence and all the blocks must be too.

Imagine using a single accounting ledger to support the financial accounts of every person and business on earth? That’s how blockchain works today.

In database land, this forces a global lock, which severely impedes performance and scaling.

In addition, you have this massive global state that which has all the smart contract data in it, and since smart contracts are composable with other smart contracts, this state must all be together. This does not easily allow for common database scaling practices like sharding.

Storage:

Blockchains used to require all data to be stored on every node. It didn’t take long for this to become a problem. The size of Ethereum data is currently over 10 TB. etherscan.io/chartsync/chai….

You can now run something called a "full node" which means you only keep the current state and a short bit of history. These are fully functional, just missing all the history. This is probably what most nodes are today. It can take hours or even days to sync up a node.

The more transactions per second (tps) you have, the worse this problem becomes. For instance, 10x the tps means roughly 10x the size and 10x the growth.

The storage problem is what makes it so your average person can’t run a node and your average company doesn’t want to deal with it. So now almost everyone accesses blockchains through a centralized API provider like Infura for instance.

Think about that for a second; you have tens of thousands of nodes eating electricity like M&Ms just so they can earn mining rewards, but nearly everybody goes through one or two companies running maybe a handful of nodes.

Security:

The core blockchain security/cryptography has been proven to be very sound. The blockchains themselves don’t get hacked. The things that do get hacked are centralized exchanges/wallets and smart contracts.

That's not saying they are bad, in fact centralized services solve a bunch of problems, just be careful which ones you use.

There is also the problem of losing your private keys (hard drive crashes, can’t remember where you put it), someone stealing your private keys, or on your passing your family not being able to access those funds… ever.

Usability:

I think this one is pretty obvious. There is not a single DApp that has a better user experience than a centralized app. Prove me wrong.

There are many, many reasons for this, I’ll just randomly list some: you have to install a chrome extension; if you’re on mobile, you have to open an app INSIDE your wallet app and doing so will break a lot of other functionality the app may have;

cost (see above); performance - it takes many seconds or minutes or hours for your action to execute; and so on.

From the developer perspective, it’s just as bad: you have to write smart contracts that can’t have any bugs; it’s extremely hard to debug; have to figure out how much gas a user will need to perform an action; ensure the user has funds; create a complicated transaction;

submit it to (most likely) a third party API provider that you probably have to use your credit card to pay for; wait for it to complete which could be a while; and so on.

Another usability issue is how things are addressed. Everything is addressed by a public key (ie: 0xBUNCHOFCHARS). So you can have 1000 token contracts with the same name/symbol and the only way to know which one is right is via...

a centralized third party, that can hopefully give you the right answer.

This is also how a lot of people get scammed, sending funds to something they think is correct, but it’s not.

Announcing Epic Ledger

Epic Ledger is a complete rethinking of blockchain and cryptocurrency platforms that aims to solve all of the above issues. It is nothing like you’ve seen before.

Epic is built from the ground up to be infinitely scalable, low cost, and simple to use.

Epic Ledger is focused entirely on asset ownership and asset transfers. Think of it like a next generation Bitcoin, not a general purpose programmable blockchain like Ethereum. It will be able to handle any number of different tokens though.

Cost:

Epic Ledger transactions will be less than a penny, period.

Energy consumption:

Epic solves this by using a smaller set of nodes. This is pretty common in newer blockchains as it's really the only way to reduce energy consumption. And there's a good argument that Bitcoin and Ethereum aren't really that decentralized anyways since...

a handful of mining companies run most of the nodes. The question is how many different organizations in different countries do you need to ensure it can’t be shut down or censored?

Scalability:

Epic Ledger solves the scalability problems by using a totally different data structure that is easily shardable and allows transactions to be processed in parallel.

Technically, it's not a blockchain (ie: a chain of blocks).

It starts with namespaces (think separate tokens) and within each namespace, each address has its own branch/ledger.

Rather than one single ledger for millions of accounts like current blockchains, Epic will have millions of ledgers.

This means we can run almost all transactions in parallel, unless two transactions are from the same namespace and account.

Epic Ledger is massively parallel and infinitely scalable.

Storage:

Epic only requires nodes to keep a link to the most recent state of an address and some small bits of metadata about namespaces and addresses. All state and history data can be stored separately in simple key/value stores such as @IPFS

Security:

Uses the same proven cryptographic technology used in existing blockchains.

Also, every wallet in Epic Ledger is a multi-signature wallet. Although, a wallet can have a single signer to act like a typical blockchain wallet. This will enable better ways to deal with things like safety, custody and beneficiaries.

Smart contracts:

How do you stop people from writing bad smart contracts? You eliminate them. Epic will not have smart contracts.

Usability:

Epic Ledger has human-friendly identifiers based on decentralized identifiers en.wikipedia.org/wiki/Decentral… which means you can send funds to an address like @elonmusk rather than some long string of characters. This will drastically reduce accidents and scams.

Similarly, namespaces provide human-friendly and much safer ways to reference tokens. You can reference tokens by name instead of by some cryptographic address, ie: USDC instead of 0x40CHARADDRESS.

Putting those together means you can safely transfer just by knowing the names like: send 10 USDC to @treeder

For developers: a nice RESTful API makes integration a piece of cake, fees can be paid by a third party and real-time responses will make for MUCH better UX.

Team:

We’re assembling an amazing team of blockchain, database, and cloud infrastructure developers with decades of experience to build this new technology.

More awesome features to watch out for: swaps and dark pools. 👀 More to come on those.

Learn more about Epic Ledger at epicledger.io and follow us @epicdlt

Now AMA!