Johannesburg, 06 May 2021
Nobody likes having the rug pulled from under them, especially if they had been expecting that rug to be around for the foreseeable future. So, when Red Hat announced the imminent end of CentOS, a widely-used version of its particular Linux operating system, the news left many upset and concerned.
It's a complicated issue infused by emotional online debate. On the one side, enthusiasts of the new CentOs Stream model list the many benefits of the new approach. They draw parallels with other enterprise-flavoured market models such as Canonical's paid support for Ubuntu, and note that - overall - this is a good direction for Red Hat. Change is good, they say, and we should embrace change.
But it's not that simple. The majority of the other side's rebuttal comes down to how the transition is being handled. Specifically, CentOS 8, the current version used by the market, will not be supported beyond the end of this year - a drastic reduction from the original timeline of 2029.
So, the debate is not just philosophically about change. It has real consequences for current CentOS users out there. One of these is Obsidian, which builds technology solutions using open source systems such as Red Hat Linux and CentOS.
"We were caught by surprise, like everyone was," says Deon Lottering, Red Hat Certified Engineer at Obsidian. "It didn't leave a warm and fuzzy feeling. But as open source ways go, we adapt and overcome."
An ability to adapt
The move raises a question: can you rely on open source if this type of thing could happen? If a vendor of proprietary software made such sudden changes to its product and lifecycle, there would be mobs in the street. Well, not mobs, but certainly groups of highly anxious and annoyed technology professionals. That guarantee of stability is, after all, part of what they pay for.
Yet, the CentOS shift actually demonstrates a value of open source that isn't often appreciated - its adaptability.
"I'm excited by the change, because this is I think the nature of open source," says Martin Liebenberg, MSP Team Lead Obsidian. "If we move beyond the shock factor, it's very exciting that a gap opened up in the market."
By this, he is referring to AlmaLinux and Rocky Linux, two alternatives to CentOS that are close to being clones (or, as Lottering refers to it, bug for bug replications). One drawback of proprietary software is the walled garden - every piece of proprietary software is virtually unique and thus near-impossible for a rival to replicate in a detailed way. Yet since they have access to the source code, open source developers can create alternatives for the market.
"That's one of the characteristics of people in open source communities," says Lottering. "We don't take offense if you improve our code. I could have done this, but you did and it was awesome."
Relying on the open source channel ecosystem
It's not all smooth sailing. Liebenberg admits that the change is an issue for service providers, "Customers rely on you to say if this will be okay or not. I think this illustrates why open source is different from other technology choices. It's a culture, not a product, and companies that use open source rely on that culture. So when this kind of thing happens, it's up to the providers who are the touchpoints of that culture, to communicate choices to customers and assume the risks of the change."
The CentOS shift demonstrates the open source dynamic in action, which can be very different from the 'nobody gets fired for buying IBM' crowd. It relies much more on community and collaboration, making open source much better at adapting to changes. Both AlmaLinux and Rocky Linux announced themselves within days of Red Hat's news, and those builds were available in four months. That is unparalleled and a feather in open source's cap.
The controversy around CentOS' changes is warranted. The announcement was abrupt and sweeping, even though Red Hat's intentions are sensible. Some have described this as an example of big business getting involved in open source, though that rhetoric would disqualify many enterprise-supported Linux builds.
It does demonstrate the uncomfortable realities of creating stability in the open source market. But this shift also demonstrates the flexibility of the market and its capacity to produce viable alternatives. It means choice, which, if managed correctly, is a considerable advantage for tech-enabled businesses.
"We might lose CentOS. But we're gaining two CentOSs," quips Lottering. "Competition is never a bad thing, especially if it can happen this quickly."
Share