How CentOS Stream and RHEL 9 Led to AlmaLinux 9

ZDNet writes that in late 2020 Red Hat decided "they'd no longer release CentOS Linux as a standalone distribution. Instead, CentOS Stream would work as a beta for RHEL."

So where are we now? The competition immediately sprang up to replace CentOS. The two most important of these are the AlmaLinux OS Foundation's AlmaLinux and Rocky Enterprise Software Foundation's Rocky Linux. [May 16th saw the release of Rocky Linux 8.6.] Now, mere weeks after the release of RHEL 9, AlmaLinux 9 has arrived.

Like RHEL itself, AlmaLinux 9 starts from CentOS Stream via RHEL. Indeed, AlmaLinux developers are CentOS Stream contributors. The bottom line is that CentOS 9 is an identical twin to RHEL 9 — except for the names and trademarks. It has all the same features, all the same advances, and, for better or worse, all the same bugs.
Besides the big server architectures, AlmaLinux is also ready to run on everything from cloud and Docker images to Microsoft's Windows Subsystem for Linux and Raspberry Pi, the article points out.

And Jack Aboutboul, AlmaLinux's Community Manager, tells ZDNet "We are building AlmaLinux with the specific goal of creating an independent CentOS successor that is truly community-centric and designed for everyone... We offer everyone a uniform platform that is safe, secure, easy to use, and dependable to build your tomorrow on."

I like to mock RMS - heck, he often deserves it

[93 Escort Wagon]

But I gotta admit the GPL is largely what makes AlmaLinux, Rocky Linux, and the like possible - despite the best efforts of Big Blue to squash them. So thank you, Richard Stallman.

Re:

[strawspulled]

Big Blue had nothing to do with any of this, you know?

Re:I like to mock RMS - heck, he often deserves it

[93 Escort Wagon]

IBM has owned Red Hat since 2019.

Re:

[_merlin]

That doesn't mea the GP is wrong. I was expecting Red Hat to kill off CentOS ever since they took control of it. You could tell they'd always been disappointed at their inability to lure CentOS users onto paid RHEL subscriptions. The only surprise is that it took them so long.

Re:

[strawspulled]

But I can tell you categorically that they had nothing to do with the decision.

Re:

[thejynxed]

Anyone who has any dealings with IBM whatsoever in the last 30 years can tell you someone at IBM did have some say in the matter because that is how IBM operates and has always operated, no matter what categorical denial you try to vocalize. They are one of the companies known world-wide for micromanagement of subsidiaries.

Re:

[ArchieBunker]

I'm still not exactly clear on why CentOS had to end.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Re:

[Anonymous Coward]

The CentOS board had been mostly hired by Red Hat before this. I'm sure that not all of the reasons are public, but more development was happening in the field with CentOS rather than RHEL, and customers would develop on CentOS and use RHEL for production. And it bumps up Red Hat's value to have licenses, even very inexpensive or free licenses, to demonstrate their presence in the world market. But various of their commercial ventures, like (not-so)FreeIPA and Ansible Tower were simply not bringing in reven

Re:

[strawspulled]

But I can tell you categorically that IBM had nothing to do with the decision.

Re:

[markdavis]

>"I'm still not exactly clear on why CentOS had to end."

It didn't have to end. IBM was under the false assumption they could force everyone using CentOS to pay up and move to RHEL or become beta testers for it instead. But it doesn't work that way in the FOSS environment, and they should have known that already. It had the exact predictable results:

1) Development of new RHEL clones.
2) People moving to other, non-RHEL derived distros.
3) Poor reputation for IBM/RedHat.

Further, the stunt severely hurt th

Re:

[mike2006]

So true. How many admins, including myself switched to Centos 8 and expecting to be good for x number of years only to be blindsided that support would be gone in a matter of months. That had to be the most stupid decision on their part and probably intentional to force purchase of subscriptions. Such an obviously shady tactic. I will have nothing to do with IBM/Redhat because of this tactic in the future and will never forget.

Re:

[strawspulled]

IBM had nothing to do with the decision.

editors?

[algaeman]

Is zdnet hiring editors from slashdot? "Indeed, AlmaLinux developers are CentOS Stream contributors. The bottom line is that CentOS 9 is an identical twin to RHEL 9"

Re:

[strawspulled]

The move to separate the Ansible core from the majority of Ansible modules was very deliberate, allowing developers of the modules to release their code at their own cadence, and not have to wait on the potentially far slower updates to the rest of Ansible. There's nothing stopping you from pinning any of your work to specific versions of collections via your requirements.yml file. IME a lot of people in the Ansible user community are actually quite happy with this change, and it made the project's mainten

Re:

[strawspulled]

How is this actually making your life more difficult beyond the time you are wasting whining about it?

Re:

[libra-dragon]

What the parent was trying to say is that the ZDNet article has an error and should probably read as: "Indeed, AlmaLinux developers are CentOS Stream contributors. The bottom line is that AlmaLinux 9 is an identical twin to RHEL 9"

Re:

[Joey Vegetables]

Minor quibble. What you are describing is not CentOS, which no longer exists, but CentOS Core.

Too Late for Me

[lsllll]

When Red Hat announced the CentOS future path, I abandoned everything Red Hat related and moved on to Debian. That's for my own stuff and the customers whose machines I control. But one of my customers decided to switch to RHEL8 and pay for the subscription. Mind you, they're a Windows shop and the Linux servers they use are pretty much just for web sites and small applications, so we will NEVER call Red Hat for support for them anyway. I could not convince them otherwise. SMH.

Re:

[account_deleted]

Comment removed based on user account deletion

Re:

[Catvid-22]

Same, switched to (the unfairly maligned) Ubuntu as it's basically Debian with layer of professional support, [...]

Debian with Snap apps? That's a big difference. IRC Ubuntu also runs a snapshot of Debian Unstable, which has naturally more bugs than Stable or Testing, and since it's a snapshot largely won't be updated within the release version. (I don't know if make exceptions for critical bugs.)

Re:

[account_deleted]

Comment removed based on user account deletion

Re:

[Antique Geekmeister]

I know of several companies who discarded RHEL and CentOS for any proof of concept projects. They consider CentOS discarding point releases that match RHEL point releases as destabilizing, though Red Hat has tried before to discard point release. The previous attempt was Red Hat 9 nearly 20 years ago in 2003. That did not work out well, either, clients stayed on Red Hat 7 point releases as long as possible, just as current Red Hat customers are remaining on RHEL 6 or RHEL 7 as long as they can.

Re:

[markdavis]

>"When Red Hat announced the CentOS future path, I abandoned everything Red Hat related and moved on to Debian. "

We were just readying a CentOS 8 major server deployment. Instead of switching and losing all that investment of time/energy, we delayed the rollout for 6 months, until a clone was released. Then ran the conversion script, and rolled it out as an AlmaLinux server instead.

Re:

[nouserbansadminonly]

Is postfix working on Debian Current yet? When CentOS died, I tried to move to Debian, but postfix was badly broken and I did not have time to port it to Debian, so I am still on CentOS 7 unfortunately. I would like to move to Debian, but as long as postfix does not run on it Debian is useless to me.

Stream is unsupportable by design

[lusid1]

The reason point releases are important, and the reason (up)Stream is unsupportable, is that you need predictable and repeatable code drops to test your products against so when you tell users an OS is 'supported' your support organization stands a chance of actually making it work. With rolling releases, there's nothing to to test against. Just because it worked on stream today is no indication that you will be able to make it work again tomorrow, so rolling releases are inherently unsupportable, and that is by design.

What we need is another Linux distro

[rantrantrant]

Let's put together a Linux distro that's outside of the control of these corporations & works as everyone knows it should. Who's game? :P

Re:

[nouserbansadminonly]

Rather sad this is marked funny. You must wait a little bit (2 minutes) before using this resource; please try again later.

Re:

[exomondo]

Rather sad this is marked funny.

None of the hundreds of distributions that already exist are good enough? Let's make another one!

Re:

[jmccue]

Well, Slackware says "Hi There". That is the Linux I use and v15 works very well.