After a few decades of covering the technology industry, déjà vu becomes an affliction one learns to live with and ignore, like a karbunkle or Donald Trump.
On Tuesday, the leading industry players in the emerging field of containerization formed a foundation. It is ostensibly for seeking solutions around how best to nurture an ecosystem around this new and fast-growing deployment system for virtual workloads.
And if you’re thinking, wait a second, Scott, you just ripped that story right out of last month and reprinted it here... no. This happens frequently in the open source software industry, the way yellow flags breed yellow flags in auto racing.
Somebody forms a foundation or a consortium, and a few weeks later, somebody does the same thing — not even necessarily somebody else.
Come Together Right Now Over Me Instead
Google is the initiator this time of the creation of the Cloud Native Computing Foundation, which will be charged with leading an industry discussion about building reference implementations for container management environments.
Google has one of the leading such environments, called Kubernetes; and even though it’s risen to prominence already in major data centers, only Tuesday Google declared Kubernetes version 1.0 generally available.
By “reference implementations,” I mean model use cases of the technology in major industries, that can be published, distributed, and implemented by companies that share the same industry space.
That’s different from a standard, which specifies how a technology or a component is constructed.
Think of a standard like a recipe, and a reference implementation like a technique a master chef uses for reproducing that recipe.
“Ultimately, it’s not about Google and it’s not about Kubernetes,” pronounced Google senior product manager Craig McLuckie — a man whose job is all about Google and all about Kubernetes — during a speech to the OSCON convention in Portland Oregon Tuesday. “It’s about bringing to market this new computing paradigm that’s going to drive radical efficiencies.
“Back in the industrial revolution,” McLuckie continued, “there was this hypothesis that, as steam engines became more efficient, coal consumption would decrease. The reality was, quite the opposite happened. As steam engines became more efficient, people found better uses for steam power. It actually drove the efficiencies and the consumption.”
McLuckie didn’t continue the analogy, leaving it for others to draw conclusions about what he really meant.
Here’s what he really meant: If your aim is to be on the receiving end of investment in the business of improving efficiency, maybe it’s better to be in the coal industry than the steam industry.
There’s Initiative - and There’s Foundation
Docker Inc. was the instigator of the Open Container Project, which was formed last month. Its stated goal at the outset was to drive the industry into forging a solid specification for what a “container” is.
In an effort to avoid the situation the virtualization industry still faces, where there is more than one format for virtual machines floating around, Docker and other industry leaders, including container maker CoreOS, agreed that everyone should follow a standard format.
But during the DockerCon conference last month, attendees — which included representatives from major containerization industry supporters — expressed their belief that a broader discussion about the container industry as a whole would emerge from this central starting point.
In my explanation of the Project last month, I said Docker Inc. faces a dilemma. For the discussion to even get started, it needed to surrender some of its claim to the ideal of containers, but not so much that it would no longer be able to lead the discussion. If Docker brings together IBM, VMware, Microsoft, Google, HP, Red Hat, and so many other names around the round table, how long before they stop considering Docker the chair of the broader discussion?
We have our answer now: about four weeks.
At the same time the Linux Foundation endorsed the creation of Google’s CNCF, the Open Container Project was given a name change... by the Linux Foundation, to the Open Container Initiative.
I’ve been told by a Docker spokesperson that nothing should be read into this.
But if you listen to the subtle focus of Linux Foundation Executive Director Jim Zemlin’s comments this morning, as to the inspiration for creating what he did acknowledge to be another consortium about what, on the surface, would appear to be the same subject, his explanation had nothing to do with the party line that the OCI is about a standard format, the CNCF is about a reference specification, and ne’er the twain shall meet.
“Let’s get into a little more nuanced view of what it means to set this up, and why it’s important for you as a developer, as a vendor, as any one participating in this ecosystem — to understand why these entities are important,” Zemlin began.
“When you invest in a new computing paradigm, when you embrace things like Kubernetes or Mesos or container technology, or any of this technology, whether you’re a company or a developer,” he proceeded, “you’re making an implicit futures contract with that technology. Is that technology going to be around five years from now? Am I going to have to pay a tax for that technology two years from now?”
Setting up a neutral entity like the CNCF, Zemlin continued, creates a protected center for the assets and other intellectual property belonging to an open source project that could come under challenge later by an outside firm.
A decade ago (which may as well be an eon ago) Microsoft threatened vendors of Linux, and products that supported Linux, with violating Microsoft intellectual property. Some of those companies entered into patent covenants with Microsoft, rather than be faced down in a courtroom by a company that spared no expenses to win there.
It was the Linux Foundation’s Jim Zemlin who threw down the first gauntlet that Microsoft would take seriously as a counter-challenge. The Foundation would serve as a neutral shelter for Linux-oriented intellectual property.
It wasn’t obvious at the time, or even several years later, but this was the beginning of the end of Microsoft’s open warfare period.
The complete transition of that company to one that would embrace the open source movement, that would participate with the Linux Foundation itself, and would directly partner with Docker Inc., began with Zemlin saying, “Enough.”
“All of these things need to be held in a neutral entity collectively supported, Zemlin continued, “so nobody can have an edge over anybody else... This is why the Cloud Native Computing Foundation has started, in order to provide that assurance to all of you as developers, or participants in the industry.”
Microsoft may never again adopt that kind of belligerent stance, but it’s not inconceivable that someone else might. Zemlin presented a picture of the CNCF as a kind of asset bank, a center of intellectual capital, backed by real funds flowing from major contributors.
If you’re potentially one of those contributors, consider this: Would you feel more comfortable seeing Docker at the chair of this asset bank that protects your interests? Or Google?
Enterprises, perhaps yours included, may invest billions in this new data center infrastructure technology over just the next few years. Docker certainly has attained a certain swagger in its two years of existence, as the acknowledged and deserving leader of the containerization space. But Google doesn’t need to swagger, and that point perhaps makes itself.