2017: When Ethereum Will Go From IT to Enterprise
Jeremy Millar began his career as one of the first Java architects at Oracle, and is now the chief instigator at blockchain firm ConsenSys, where he plays a lead in its evolving enterprise strategy.
In this CoinDesk 2016 in Review special feature, Millar discusses how he believes the community galvanizing around the ethereum blockchain will continue to grow – and gain big business support – in 2017.
Enterprise use of blockchain technology has evolved at an almost unfathomable rate over the past 24 months.
From early bitcoin experiments, to senior bankers joining startups, to the launch of the public ethereum decentralized application platform, to the many private, permissioned systems using the technology, blockchain has emerged as one of the top enterprise IT trends entering 2017.
Yet, the market has already moved beyond the incubation phase where innovators effectively build the technology along with their initial applications, and possibly beyond the early adopter phase, too.
Increasingly, mainstream enterprise IT organizations are not only educating themselves and experimenting with blockchain, they are also aiming to tackle novel use cases and complex IT challenges with the technology.
More and more frequently, our clients are asking for assistance building MVPs not PoCs, or hardening environments for production readiness.
With this whirlwind of adoption, it is also clear that certain key technologies are emerging as potential de facto standards as blockchain platforms.
Why IT loves ethereum
Ethereum is, arguably, the most commonly used blockchain technology for enterprise development today.
With more than 20,000 developers globally, the benefits of a public chain holding roughly $1bn of value, not to mention an emerging open-source ecosystem of development tools, it is little wonder that Accenture observed "every self-respecting innovation lab" is running and experimenting with it.
Cloud vendors are also supporting ethereum as a first-class citizen: Alibaba Cloud, Microsoft Azure, RedHat OpenShift, Pivotal CloudFoundry all feature ethereum as one of their, if not their primary, blockchain offering.
Why? The software is widely available and its simple to download an ethereum client; pick your favorite development environment and get going.
Ethereum is general purpose and easy to program – full stack and web developers can pick up the basics of the Solidity smart contract programming language in a matter of hours and develop initial applications in days.
Documentation is plentiful, as are code samples, deployment frameworks and training. Little wonder that so many companies are using ethereum as their blockchain of choice.
Today, enterprises are deploying private ethereum networks in or near production in areas as diverse as supply chain tracking, payments, data privacy, compliance and asset tokenization just to name a few.
Three challenges
Now certainly, we are some time away from seeing investment banks fully migrate securities clearing and settlement to ethereum networks.
That said, however, we already do see private ethereum blockchain networks in production – even in financial services.
Enterprises adopting ethereum face a number of challenges, notably:
Ethereum was developed initially for public chain deployment, where trustless transaction requirements outweigh absolute performance. The current public chain consensus algorithms (notably, proof of work) are overkill for networks with trusted actors and high throughput requirements.
Public chains by definition have limited (at least initially) privacy and permissioning requirements. Although ethereum does enable permissioning to be implemented within the smart contract and network layers, it is not readily compatible 'out of the box' with traditional enterprise security and identity architectures, or data privacy requirements.
Naturally, the current Ethereum Improvement Process is largely dominated by public chain matters, and it has been challenging for enterprise IT requirements to be prioritized within it.
As a result, many enterprises who have implemented private ethereum networks have either 'tweaked' or forked open-source implementations, or relied on proprietary vendor extensions to meet their requirements.
Some of these are extremely sophisticated and are on the cutting edge of computer science; for example: BlockApps STRATO, Hydrachain, Quorum, Parity, Dfinity and Raiden.
While understandable, and in fact until now, the only effective approach, the downsides are obvious: lack of application portability, code base fragmentation, and vendor lock-in.
Historical parallels
Not surprisingly, this has been a point of conversation for some months between enterprise tech vendors, corporate users and ethereum startups.
These discussions have expanded, with the blessing and involvement of ethereum creator Vitalik Buterin and the non-profit Ethereum Foundation, into a dedicated group of vendors where the largest corporate users and ethereum infrastructure leaders are collaborating to define a roadmap, legal structure, governance and initial technical developments.
To some extent, this parallels the paths of other significant platform technologies, such as TCP/IP and HTTP and perhaps (from a software perspective) more relevantly Java and Hadoop.
Java was never intended to be a broadly used enterprise development tool; it was in fact developed originally for interactive television (specifically set-top boxes and smart cards – who remembers Java Card?).
However, Java had many advantages for web development with database back-ends (known as web client-server or three-tier architecture): it had comprehensive web and database APIs, it provided 'write-once, run anywhere' platform portability, simplified object-oriented programming constructs with familiar syntax, and a rapidly development ecosystem.
Indeed, it was not even Sun that created Java Enterprise Edition (at that time, J2EE); it was a plucky startup (WebLogic) and a group of enterprise customers and other vendors. Similarly, Hadoop was originally created to index the web and for advertising serving.
And who knew TCP/IP would emerge into a protocol that today, literally exists everywhere?
Solutions on the way
In this way, I would argue ethereum is one of the few, indeed perhaps the only, blockchain technology with a similar trajectory and potential.
Even the early "public permissionless" vs "private permissioned" differentiation strongly echoes the Internet vs intranet deployment considerations that were so prevalent before businesses became comfortable with security and scalability issues of public infrastructure.
By banding together the key adopters, supporters and shapers of enterprise usage of ethereum, we are seeking to provide a platform not only for the technology, but also to provide the governance and tools to create a standard for 'Enterprise Ethereum'.
It is a group of builders and doers, developing sufficient governance for enterprise requirements, but not 'death by committee' and without 'pay to play'.
Some of our collaborators have noted the refreshing nature of this approach and the pace of technical progress that is achievable from working off a single standard and open-source code base.
Moreover, Enterprise Ethereum will build upon the current ethereum scaling roadmap and maintain compatibility and interoperability with public ethereum. In fact, we believe Enterprise Ethereum will soon contribute significantly to the overall development of ethereum.