Taking a bootstrapped initiative to a healthy open-source project is difficult. But when there’s only approximately 100 developers in the world that have a deep understanding of the technology, such as blockchain, the difficulty increases dramatically.
Open-source veteran Brian Behlendorf was aware of the challenges when the Linux Foundation tapped him to lead the Hyperledger Project as its executive director in May.
“The job really is to be an independent voice for the project that is not affiliated with one company or another,” he told Markets Media. “It’s also to bring to the party everything that the Linux Foundation knows about running open-source projects. My job is to corral all that towards the purpose of building a great community and a great collection of code.”
Behlendorf has watched the development of digital currencies, such as Bitcoin and Ethereum, but has been somewhat skeptical about some of the more fantastic claims that have been made about the technology.
Brian Behlendorf,
Linux Foundation
“This class of technology solves a problem that no one else really understood how to solve: How to truly trust a network and not just a central party in that network,” he said. “Everything else reduces to minimal viable centralization, like DNS or how we assign IP addresses. All of that requires some sort of coordinating body. With blockchain, as long as you have enough participation on the network, users can trust the network.”
The Hyperledger Project, which only formed in February with 30 members, started with a solid set of code as IBM donated its OpenChain code base and a technical steering committee already in place.
“It felt like there was a lot of substance to this,” said Behlendorf.
In any healthy and sustainable open-source project there is a staircase of participant types, including novice users; expert users who helps others as well as finds and reports bugs; and core contributors.
“A good open-source project is one where unnecessary obstacles to climbing from one step to another are minimal,” he said.
Behlendorf estimates that there are only about 100 people globally with deep understanding of blockchain architecture and another 100 or 200 people familiar with the technology. He’s spending most of his time in outreach.
He expects the number of core contributor to grow naturally as more companies are spending more money on blockchain developers and spinning up their recruitment processes.
However, he would like to see the middle tier of expert users grow as well so more people can help develop a common blockchain vocabulary.
“We use terms like blockchain and distributed ledger interchangeably and that can be confusing to ourselves, let alone other people,” Behlendorf said.
The Hyperledger Project is not interested in developing one blockchain to rule them all, but an architecture that will support a variety of blockchain use cases. A blockchain’s use case also will drive a different set of needs for those blockchains, such as which consensus mechanism or smart contract platform they should use.
Use cases can be separated into permissioned and non-permissioned blockchains. Permissioned chains are ones where the blockchain participants are known to other users and need permission to join the blockchain. Non-permissioned blockchains, like those underpinning digital currencies like Bitcoin and Ethereum, do not have such requirements.
“We’re not trying to push one particular chain,” said Behlendorf. “We will have the flexibility to evolve the platform and integrate new features without having to worry directly about how you change the engines on a jet liner in mid-flight, which is the case for some of the public currency projects. Most of the private chains will have the luxury of taking the system down for five minutes, swap things out, and turn it back on.”
It’s too early to tell how much of the differences between permissioned and non-permissioned blockchains can be abstracted away, he added. “I’m trying to keep the decision very organic and technologist-driven rather than a political decision or a mandate from on high.”
Similar to the Linux OS kernel, Behlendorf sees Hyperledger Project developing a pluggable architecture so that the same code base could support very different types of blockchains.
“I don’t think that Linus Torvalds knew, anticipated or planned for the idea that the code he released in 1991 would eventually run on everything from a Raspberry Pi to an IBM mainframe and from drones to the International Space Station,” he said. “All of these scenarios run the Linux kernel because of its intense modularization.”
Since Linux’s process management, file-system management, and almost everything else operate as modules, developer can dial up or down their distribution’s core backbone. he added.
Besides making the Hyperledger project enticing and as user friendly as possible to potential participants and gathering additional use cases, he would also would like to see the project make a couple of deliveries by the end of the year.
“We have a code base that is functional in a demo environment,” he said. “You can start to play with it and deploy prototype chains. It’s not quite at the pilot friendly level yet.”
The project should be capable of getting a firm concept of use cases for a version 1.0 release by that time and ship something in a number of single-digit months, he added. “Success to me, would be having code that at the very least is runnable in a prototype or proof-of-concept mode not just for the people who have been working on the code base, but for developers and organizations who are new to the concept of distributed ledgers.”
More on Hyperledger: