2021, for me, was a year overwhelmingly about one blog post.
On December 8 2021, I posted this blog post:
https://blog.centos.org/2020/12/future-is-centos-stream/ – CentOS Project shifts focus to CentOS Stream
This ignited a firestorm that has raged all year, with articles almost every week about what we could have, should have, or might have done differently, and what the fallout will be for the coming years, for Linux, Red Hat, IBM, and technology in general.
There have, of course, been a LOT of tech controversies this year, but this is the one that I was responsible for.
Explaining my 2021 to people is complicated. A thing that people liked was changed in a way too nuanced to explain without a history lesson going back to 2004 and people were angry and it was my job to make it better. Here's hoping 2022 is better in every way.
— Rich Bowen (@rbowen) December 31, 2021
CentOS Linux was a thing that a lot of people cared about. We changed it with an announcement of a new model, back in September of 2019. However, due to … reasons … the majority of the world missed the implications of this announcement until it was shoved in their faces in December of 2020.
That is, of course, largely on me, as it was my job to promote this new model. It is also, of course, on other people, since the notion that we would eventually switch entirely from the old model to the new model, while COMPLETELY obvious to anyone paying attention, wasn’t obvious to the majority of the user community, who had other things to be worried about.
2021, for me, was about rebuilding trust. The Dec 8 2020 announcement was, like I said, the obvious outcome of the September 2019 announcement. But when you give people something for free for 15 years, it’s reasonable for them to assume that you’ll keep doing it for another 15 years. We definitely could have handled that better. But, as Steven King says, the past is obdurate.
Today – 2021-12-31 – is the final day of CentOS Linux 8. And, frankly, the time had come for a new model. I think everyone is in (almost) universal agreement that CentOS Stream is a good thing. This was never about that. It was about how we messaged it.
Surprise, the wisdom goes, is the opposite of engagement. We surprised the user community.
It’s fine to say that they *shouldn’t* have been surprised. Y’know, if they had been paying attention. But we spent 17 years training them *not* to pay attention. CentOS Linux was thus, and would always be thus, we implied. Never, ever, promised. Never stated. Never guaranteed. But implied. And so people built their businesses – their livelihood – around that implied promise.
So, sure, we never promised. We always were very clear that we lived and died at the whim of RHEL engineering. But, of course, to some people, that was, in fact, the promise – that we would forever live in that space where we rode on the tailcoats of RHEL.
And, so, as we go into 2022, I’m definitely proud of some of the things we’ve accomplished in 2021. We have moved from being just a clone of work that others do, to being a community that encourages participation.
It’s important to note, there have always been a core of passionate, dedicated contributors, but that they couldn’t change the base platform without compromising the definition of the project as a 1-1 clone of RHEL. Those people kept the project going for 17 years, and I don’t want, in any way, to downplay their contribution.
We have added several important new SIGs in 2021 – Automotive, Infra, KMods, Documentation, and Hyperscale.
And we have rewritten the governance of the project so that I can, today, really say that this is an open project. This is the thing that I’m most proud of.
And, again, I don’t want to downplay the work of the CentOS Core team for 15+ years – they were doing awesome work, in their weekends-and-evenings, without any financial support. But they were, for the most part, doing so without any real input from the larger community. Not because they didn’t care, but because their mission – produce an exact replica of the RHEL distribution – didn’t require community input in any meaningful way.
In 2021, we have made the board process more open (public board meetings and an open nomination process for board membership) and made the board less Red Hat centric. This is, probably, my biggest accomplishment in the past year. And I’m hugely proud of that, and appreciate the help of the entire board in that process, especially Jim Perrin, Brian Exelbierd, Carl George, and Karanbir Singh.
Edit/Correction 2022-01-31: Carl is not a Board member, but has been very active in crafting and disseminating the message around CentOS Stream.
We also saw the rise of the “competing” RHEL rebuilds, and, frankly, I think this is a net win for the Linux ecosystem as a whole. Between Alma and Rocky, the “we build stuff on top of RHEL” ecosystem is stronger than it has ever been. Alma has been, for the most part, cooperative and friendly, while Rocky has been built mainly on a platform of “Red Hat betrayed us and WE WILL HAVE OUR REVENGE”, but I am encouraged about how this conversation will become more productive in 2022.
Of course, I must mention Jack, who has been a breath of fresh air this year. Sometimes you encounter people who are too genuine, too sincere, and you question whether it’s for real? Jack is one of those people. And over the course of 2021, it has been clear that he is for real. He cares about community. He cares about the people that work and live and depend on our technologies. He is genuine. And, let’s be frank here – that’s not common.
And, so, I say goodbye to the most difficult, stressful year of my professional career (at least, I certainly hope so), and look forward to the coming year. I think we’ve done good things in this difficult year, and that it will bear fruit in the coming year.