Crypto — RadixDLT: A dose of nuance: Scalability (millions and millions of TPS!?)
This piece arose out of a bit of irritation with the widespread tactic of creating hype, even FOMO with superficial claims in the crypto space.
Of course, marketing usually is that shiny piece of silver that attracts birds that are flying by. Therefore it shies away from getting into the nuts and bolts of the good or service they try to gain awareness for. The technicals are often longwinded or to be honest boring to most customers. But when every project represents itself as a piece of silver the entire space becomes blinding with all the light it is reflecting. Let’s land and inspect if there really is value.
Scalability is one of those topics that gets abused by claims of millions and millions of TPS. I would like to inject some nuance in this area by pointing out that there are different types of scalability. Consequently, scalability statements should specify to which type it refers to.
Let’s enter deep shard space!
There are 2 categories of transactions to be discerned on a distributed ledger. The distinction is made by the presence of a smart contract. Application logic scaling is on most DLTs much harder to do than simple value transfers. This is also the case for Ethereum as can be seen on below graphic used by Vitalik Buterin on a presentation given on Coindesks virtual conference ‘CoinDesk invest: ethereum economy’ that started on 14th of October 2020.
‘Generic EVM applications’ are trailing behind the ‘simple payments’
In those 2 categories there are 3 distinct types of transactions
Visual representations can be found in below images. The article header image provides an explanation of all the components we can see here (check it out first before continuing here as things will be much easier to follow).
These types are ordered from least complex to most complex.
Although this is an abstraction for the purpose of conceptually being able to talk about this subject matter. How it technically looks like might be different from ledger to ledger. This because the answer might be different on the question: ‘What is being stored on each shard?’
Let’s couple back to Radix.
According to Dan Hughes, founder and CTO of Radix, both ‘simple payments’ and ‘generic EVM applications’ in the terms of Vitalik will scale to the same degree, and that is linearly. Below an excerpt of the AMA session held on Radix’ telegram channel
Every loved pet has a nickname. ‘Cerby’ is the nickname of ‘Cerberus’ which is the name for the novel distributed ledger technology Radix research and developed over the course of 7 years. Many iterations preceded it, beginning with the well-known blockchain. I would recommend this article to continue your rabbit hole journey into Radix: https://medium.com/@radixdlt/dan-and-radixs-tech-journey-70752de17629
Here the same concern is uttered. TPS claims are often only applicable for the simple value transfers which are the easiest to scale. The real proof is in the pudding called smart contract scaling.
I’ve started off with a sharded ledger representation to set myself up for a little bridge into ‘Composability’ (to be expanded on in a future article).
I want to put out there that the blockchain trilemma (scalability, decentralization, security) coined by Vitalik might not be telling the complete story. According to Radix there’s a fourth element namely composability.
When sharding is implemented in favour of scalability, there is a risk elements are broken in this new quadrilemma (scalability, decentralization, security, composability)!
When dapps are located in different shards of a blockchain they are unable to be atomically composed together. This means that different smart contract transactions of these dapps can not be put together as one anymore because they reside on different shards. They have to communicate via an intermediary (often a chain) which introduces latency and increases complexity. This would mean a whole array of (combined) functionalities dapps provide on an unsharded ledger are broken (e.g. flash loans). I do want to include that grouping related dapps in shards is a workaround but then there might creep in once again scalability issues and decentralization issues.
Radix claims to have come up with a solution that keeps the elements of the quadrilemma intact. This intershard case is resolved by a consensus mechanism they named ‘Braiding’ (this as well to be expaned upon in a future article together with ‘Composability’).
What I’d like you to takeway from this article are 2 key messages:
- There are different types of transactions to be scaled with each their own difficulty.
=> When projects claim a number of TPS ask yourself: “Are these claimed TPS numbers applicable on all types of transactions?” - With the introduction of sharding there is an element that was taken for granted namely ‘composability’. Sharding a blockchain made it come to the forefront as the fourth piece of the puzzle.
=> When projects claim high scalability ask yourself: “Is scalability achieved while keeping composability (and decentralization and security) intact?”
How important composability is depends on the type of application but it seems like that the strength of most dapps comes from this tying together of multiple features as one. Succes as one, failure as one. Alot of these dapps enable functionalities that weren’t before possible. Functionalities in both the individual as well as composed (combined as one) sense.
If you are a crypto enthusiast (for whatever reason: idealistic, technological, investment wise, …) just like I am then I would recommend looking into what Radix has to offer. On all fronts this project is very exciting. I sound like a shill right now so the best course of action is to perform your own trial by critical fire of Radix’ claims. Your motives can be to gain extra insight in the value of the crypto projects you are currently invested into or to scout for the next opportunity.
Interested to know more? Want to join the discussion? Or ready to sign up as a decentralized application developer on top of Radix? Follow the breadcrumbs below:
Dev sign up form: https://radixdlt.typeform.com/to/JHBeokkZ
Telegram: https://t.me/radix_dlt
Website: https://www.radixdlt.com/
Twitter: https://twitter.com/radixdlt