Ripple's former CTO Stefan Thomas is going up against Ethereum with the launch of a new smart contracts platform.
Well, «new» isn't quite right – the platform Thomas released implementations for today is Codius, an open-source project that Ripple released in beta in 2014 but shelved the following year. Now, though, having announced his departure from Ripple in May, Thomas is re-launching Codius as the technical backbone for his new company, Coil.
Using Codius, Coil aims to change the way websites monetize their content.
According to Thomas, monetizing web content has so far relied on clunky «workarounds» like ads, paywalls and user data harvesting (think Facebook's recent debacle). But his new project, by using Interledger, an open-source protocol that was developed inside Ripple for sending payments across different ledgers, plans to allow users' browsers to make micropayments to websites they visit.
Codius could enable use cases such as a «revenue disbursement contract,» which could take in revenue as people watch a movie and pay that money out to all the parties that made the movie – and not in batch payments, but little by little. Or a Codius smart contract could help news outlets and their readers interact in that it could manage readers' authorizations and subscriptions «and act as a sort of switching board for your money,» Thomas said.
The implementation released today comes with tutorials for uploading and hosting Codius smart contracts (uploaders pay hosts to run smart contracts on their computers), to try and push developers to start using the platform right away.
And already several developers have disclosed they'll be building on the platform, as revealed exclusively.
Telindus, an Luxembourg-based IT solutions subsidiary of the state-owned Belgian telecom Proximus Group, will use Codius to «push forward novel direct e-commerce models,» Telindus chief architect Thomas Scherer.
Josh Williams – who has previously invested in well-known game platforms Unity, Zynga and Kabam – said he'd be using Codius in new ventures, including a gaming company that is currently in stealth.
«Teams in games and elsewhere are building on Ethereum and running into the cost and scalability issues we're all familiar with. Codius has great potential in addressing these concerns, and we are eager to work with it.»
Thomas echoed that sentiment, saying that as Ethereum has demonstrated the viability of smart contract use cases, it has simultaneously shown the world its own vulnerabilities, as Ethereum-based applications continue to run into scaling difficulties.
In contrast to Ethereum, Codius was designed to allow developers to write smart contract code in any programming language and have the smart contracts work as «smart oracles,» communicating with outside data sources.
As such, Thomas said, Codius has an opening.
«The people that are reaching out to us are saying, 'Hey, we're experimenting on Ethereum. We're running into scalability issues. It's too expensive, too slow. It's not flexible enough. We don't like writing in this awkward language,'» he said.
So why did Ripple set Codius aside?
While Codius generated its share of buzz in early 2015, before Ethereum's mainnet was live, according to Thomas, the idea seemed premature. Ripple engineers touted the platform as a model for interoperability at the time, saying it was able to handle not just XRP – the cryptocurrency most closely associated with Ripple – but bitcoin, ether and fiat currencies.
But the project hit snags, however.
Adding smart contracts opened up new ways to attack the ledger, and the technical architecture was cumbersome. Speaking in 2015, Thomas said that building smart contracts into a blockchain was like writing software directly into a database – difficult.
The team realized, in a recent interview, that computer science had solved that problem in the 1970s by developing a three-tier architecture, in which a «logic layer» sits between the database and user interface layers.
Codius would serve as that middle layer, Thomas said, adding, «You would have a bit of code that's accessing some assets on XRP ledger, that's accessing some data that's in Ethereum, and maybe it's making an HTTP call and so you have a much more flexible architecture. And most importantly you can have those kinds of contracts call other contracts as well.»
But building that kind of platform required efficient communication between ledgers, something that wasn't available at the time, and so Ripple began developing the open-source Interledger Protocol to allow for this communication.
Plus, Thomas said:
«We just didn't feel like smart contracts was a very mature industry at that point…. Frankly, the use cases seemed somewhat dubious in value.»
As such, Codius was shelved. But now, three years later, Thomas' doubts about the value of such a smart contract platform have disappeared.
Rather, he sees Ethereum's scaling issues – expensive transactions and slow confirmation times – as signs that smart contracts are ready «to move away from the mainframes, move away from Ethereum and go over to a more flexible architecture that involves multiple different ledgers.»
Ethereum killer?Codius, though, is hardly the only would-be Ethereum-killer to emerge over the past couple years.
All of these projects, like Codius, tout the ability to process faster and cheaper transactions, but there's typically a trade-off involved – either in terms of security or blockchain's defining benefit, decentralization.
EOS, for instance, a delegated proof-of-stake cryptocurrency project which is currently in the midst of launching on mainnet, promises faster, cheaper transactions, because the blockchain only needs to be verified by 21 validator nodes, not by the whole distributed community of miners as in Ethereum and bitcoin.
Thomas, though, argues that Codius' design enables developers to balance their own priorities, rather than having to accept the network's compromises as a given.
«You can choose the level of decentralization,» Thomas told. «If you upload it to four or five hosts, you'll have a decentralization level that's similar to Ethereum [and] you'll have a cost that's still orders of magnitude lower. Or you can upload it to 100 hosts, and you'll have a much greater level of decentralization than you can get with Ethereum.»
As it relates to security, Thomas argues, Codius has several advantages over Ethereum and other smart contract blockchains.
«I think that a lot of the issues and compromises, big hacks and so on have been directly related to the fact that these are all new languages whose security's not very well understood,» said Thomas.
As for price, Thomas contrasts Ethereum's transaction costs – which can exceed 60 cents and even a dollar – with those of Amazon Web Services' Lambda platform, which costs 20 cents per million requests. AWS is centralized, but Thomas still expects Codius' costs to fall «between those two extremes.»
For Thomas, rolling out Codius is the first step towards building a standard protocol for monetizing web content, as well as the ecosystem around it. Eventually, Thomas believes firms might decide to host websites on Codius rather than AWS, with Coil serving as a kind of «Spotify, but in an open way» – a protocol connecting consumers, internet service providers, websites and content creators.
And while the code is still «rough,» Thomas remains optimistic that Codius will prove a step forward for developers writing smart contracts that solve problems for today's businesses.
«From a cost, scalability and security standpoint, as well as the flexibility … it's orders of magnitude more viable for mainstream use cases.»Correction: An earlier version of this article stated that Josh Williams worked at Unity, Zynga and Kabam. He was an investor in those companies.