Whether it’s called decentralised or distributed, the future of the internet is fundamentally different — both closer to its roots, and further away. The genesis of the internet was decentralised, but it was also inefficient — machines communicated with machines peer to peer but didn’t function so well as the sort of network the internet was becoming. It was more technical, less intuitive, and could be frustratingly awkward when hosts went down or connections failed.
This all began to change when Tim Berners-Lee developed HyperText Transfer Protocol and with it accidentally birthed what became the web and the modern internet. As using the internet moved to the mainstream the need for compute and storage grew with it, and with that, the notion of a server farm. This was when things started to get a little complicated.
Farm life is serious business
Prior to the web, most of the computing and storage was done locally, on the user’s machine. In the web model, most of the compute and storage is done by the server and transmitted to the user via the internet. This continually increasing demand for server power led to businesses investing in first their own server, then their own server farm, and for some of the larger players, their own data centres.
Of course, setting up and running a data centre is a massive risk — it’s expensive to build and run, requiring massive cooling and enough electricity to power a small town — and if it’s too small? Build another one! Too big? Wasted resources. And if something happens to it? Well, then your internet presence disappears, and your business (and reputation!) with it. So, better to build two for redundancy.
Eventually, this led to some businesses renting out their server space, and cloud computing was born — along with the behemoth that is AWS. While this did make the web grow faster and deliver richer content, it also put what some saw as a dangerously large amount of the internet under the control of just 5 companies, with 79.4% of cloud usage controlled by AWS, Microsoft, Google, Rackspace and IBM. This started some bright software engineers looking for an alternative.
Working down the mine
After the Bitcoin protocol introduced blockchain to the world at large in 2009, many of these people realised it could be the key to decentralising the internet (again) by providing the layer of security and transparency needed. The first real contender on this front was Ethereum, founded in 2015 by nineteen year-old Vitalik Buterin. Ethereum was designed to allow the deployment of decentralised applications, or dApps, as the first step to decentralising the internet, but has never realised viable scalability. Amongst other reasons, this is because the nodes that make it work are compensated by competitive mining.
Without going into too much detail, mining involves solving a complex cryptographic problem as part of the validating process for achieving consensus and registering a new block on the blockchain. Unfortunately, this process is both slow and competitive, meaning increasingly large amounts of compute are used by many participants to process a single transaction, and only one participant is compensated by the single reward of ethers. While this functions for the purposes of cryptocurrency transactions, it is not a fast process, making decentralised networking nigh impossible. Even Vitalik himself has acknowledged the shortcomings, saying “if you want to build a decentralised Uber and Lyft on top of an unscalable Ethereum, you are screwed. Full stop.”
Thankfully for the decentralists, Ethereum and its derivatives were a beginning, not an end. By learning from its shortcomings and addressing them through entirely new code, Catalyst has created a complete decentralised operating system stack that is both compatible with existing infrastructure and massively scalable and fast. This has been achieved through a novel form of consensus and distributed processing that is collaborative, instead of competitive — meaning that all participants are compensated for their work — and extremely light — meaning that nodes can be run on practically anything, including common IoT devices.
The unprecedented ease of node deployment, software inter-compatibility and the egalitarian reward structure makes Catalyst the first of a new generation of decentralised networks — one in which functionality and ease of use replace short-sighted altcoin speculation, and robust, enterprise-grade applications are easily possible. With Catalyst, there’s no need to sacrifice the convenience of the cloud to achieve the resilience of decentralisation — on a better network, we can have both.