Surely the biggest news in the open source community this year was the official announcement of the Red Hat acquisition by IBM, passing any potential regulatory hurdles, bringing one of the most influential OS companies into the fold of Big Blue.
Commentators have talked of open source going 'mainstream' since Microsoft CEO Satya Nadella's public embrace of community-led technologies, but truly, a deal of this size by one of the oldest big tech players in the industry? It doesn't get much more mainstream than that.
But where next in the landscape of open source? Well, among other things, we expect businesses to continue to experiment with Kubernetes, the formation of another Foundation or two, major leaps in open source hardware, and continued uptick in activity from emerging global powerhouses such as China. Read on for more...
Container? I barely knew her
Look – containers are hardly new – in fact Aquasec has published a potted history of the independently deployable packages of code from the '70s through to 2017, and it's worth a read.
What has captivated businesses lately is the container orchestration platform Kubernetes, which was let loose from Google's labs into the world in 2015. According to the Cloud Native Computing Foundation (CNCF), 40 percent of 5,000 enterprise businesses that it surveyed are running Kubernetes in production, cementing its place as one of the de facto darlings of the open source world.
But, as we pointed out on sister title InfoWorld, those companies that are running Kubernetes in production are doing so on a small scale. The platform is famously difficult to grapple with, and no end of workshops and panels at open source conferences have combed over the finer details, the pros and cons, and shared their experiences with how to run it successfully.
So the race now, at least in the open source vendor space, seems to be monetising the orchestration platform, with big hitters like AWS, Cisco, Microsoft, Pivotal, IBM-Red Hat, Google and HPE all launching their own managed versions. Customers can see the benefits but they're struggling with completing the work at the coal-face – a sizeable time and engineering commitment.
But that's not to say that managed services from the major cloud vendors will be the only way forward. Cash-, time- and engineer-rich organisations with the inclination to throw resources at the problem probably will, and the global community is so active that improvements around that complexity issue will steadily come into focus.
Going into 2020, Abby Kearns, executive director of the Cloud Foundry Foundation, says that a move towards simplification will be key: "2020 will be about simplifying abstraction development and deployment," she said. "This year, we saw developers turn to technologies that abstract away the complexity and provide consistency across platforms as they are faced with an increasingly complex multi-cloud environment."
However, Patrick McFadin, VP for developer relations at DataStax, believes that there will be a certain level of "disillusionment" in the coming year as the hype surrounding Kubernetes starts to fade. McFadin predicts difficulties surrounding building and maintaining real-world deployments may take the shine off somewhat, however, this should be understood as a "good sign of a strong future" and typical of lifecycle patterns.
"As a result, you will see a lot more antipattern talks," said McFadin. "New projects trying to take its place with improvements, voices of doubt at why they ever started a project using Kubernetes... In all, the community of operators considering Kubernetes for massive deployments will use more caution and careful evaluation."
The question isn't will this or that leading vendor buy out active open source communities, but more: why wouldn't they? Businesses that are increasingly having to rely on open source technologies across their stack gain cultural, technological, and development cache, provided they're sufficiently hands-off and don't asset-strip (or micromanage) these acquisitions to death.
We'll not pin a guess on who might be headed to the auction-house. Melissa Di Donato just promised us that SAP won't buy SUSE, though.
Cloud Foundry's Abby Kearns suggested that if not M&As, there will be some form of consolidation on the way next year: "We predicted 2019 would be the year of consolidation among major enterprise tech companies itching to get a piece of the latest innovations.
"Indeed, we saw some major mergers and acquisitions this year, and the market is continuing to consolidate around specific technologies. In 2020, we believe those technologies are Kubernetes and, more broadly, cloud. Organisations are consolidating around their cloud footprint and while many will continue to employ a multi-cloud strategy, the major players will make themselves known."
Open source hardware
Although floated for a while as an intriguing fork into the future of open source, there have been signs throughout 2019 that strides in open source hardware – specifically chip design – could well be on the cusp.
Late in 2019, Google and Cambridge-based non-profit LowRisc announced their progress with an open source root of trust system-on-chip blueprint called OpenTitan. The root of trust layer of hardware security is designed to prove that a machine has been free from tampering as well as allowing it to verify software updates and roll back to earlier versions, the Register explains. But if microprocessors themselves can't be trusted? By open sourcing a design for the silicon, it would be possible to totally scrutinise the security of the system.
The RISC-V Foundation is packed with major technology players, but as TechNode points out, the Foundation itself also has garnered interest from the biggest Chinese technology companies. China has always been open about moving away from its reliance on American technology, but with the Trump-led trade war there is, no doubt, even greater impetus today. The RISC-V project could get the ball rolling.
Elsewhere, Alibaba's chip arm Pingtouge open sourced a microcontroller/MCU design platform on 21 October. It uses the RISC-V architecture Xuantie 902 processor, and Alibaba hopes that the project would be of interest to chip design companies as well as universities and research institutes, particularly in the sphere of AI research.
And the October edition of the Economist stated honestly that while open source might be a bewildering concept to your average closed-off capitalist, it charted the potential for growth (as well as the underlying political reasons – more on that in a second) in open source hardware, also mentioning, of course, IBM open sourcing its Power chip designs.
(Trump voice) CHINA
The Trump-China trade war and America's targeting of Huawei has created a technology rift between the two countries that has ramifications that could be far greater than China's desire to isolate its online, digital economy.
For a rising superpower, reliance on hardware and software from a potentially hostile nation is simply untenable – this is not paranoia but has great precedence, what with the internet's secret history as a tool for spooks and spies, not to mention the Five Eyes arrangement between Britain, the USA, New Zealanda, Canada, and Australia, revealed in depth by whistleblower Edward Snowden to be a gigantic snooping dragnet.
China might be the most glaring recent example, but sanctions have impacted software delivery in other parts of the world too – with Microsoft distributors halting sales to organisations in Russia, Iranian accounts being removed from Github, and for a time, Adobe limiting its delivery of Creative Cloud to Venezuela. (More on this to come from us in the new year.)
So there is already an appetite for alternative models of infrastructure and software building – just as there are early inclinations of an appetite for an alternative mode of digital economy, modelled on China's rather than America's – but there are early signs that this may grow as hostilities deepen.
Open source will be absolutely integral here. As Computerworld has stressed for years, organisations within China much prefer open models because they have complete visibility into the stack, and a peek at the big OS projects reveals how valuable upstream contributions coming from China are, too.
Software is eating the world – and open source could either prove fertile battleground or a unifying factor. In a sense, there's a lot about open source that highlights the general contradiction of a globalised economy that has hostile states operating within it. See also: our feature about the American military industrial complex's long associations with open source.
FOSS absolutely cuts across borders, and the biggest projects are world-wide phenomenon – so it'll be interesting to witness how the political events develop within the movement at large.