'What Happened with CentOS Will Not Happen with Rocky Linux' (itworldcanada.com) 47
Here's a Linux distro scoop from IT World Canada. "Gregory Kurtzer, who founded and once led the former open-source project CentOS Linux as well as The cAos Foundation, the organization where early development of it took place, said today a governance structure has been put in place that will keep Rocky Linux in the public domain forever."
Development of Rocky Linux began shortly after, in late 2020, Red Hat terminated development of CentOS, a community-based Linux distribution derived from Red Hat Enterprise Linux (RHEL) that had been in existence since 2004. It is named after Jason Dale "Rocky" McGaugh, a talented programmer involved in CentOS development, who passed away in December 2004 at the age of only 30. Asked what McGaugh might have thought of the OS being named after him, Kurtzer told IT World Canada, "to be honest, he was a shy guy. I don't know if he would have liked the attention, but at the same token, he was a huge advocate of open source and a big fan of open source.
"Personally, I don't think he would have liked what happened with CentOS."
Kurtzer added that "what we are doing with Rocky Linux is really where he would have liked to see the project and open source going. When we named it Rocky Linux, it was a hat tip to him for everything he has done, not only in open source and high-performance computing (HPC), but also with the CentOS project.
"One of the last e-mails that he wrote to the e-mail list was that he was 99 per cent done development of CentOS. It was pretty much ready to go when he passed, but he never saw it released."
The key for an open-source initiative to grow and flourish, said Kurtzer, lies with registering it as a non-profit organization, which was the case with The cAos Foundation. He has done the same with Rocky Linux.
It's official name is the Rocky Enterprise Software Foundation, "backed by an advisory board of trusted individuals and team leads from the Rocky Linux community."
"Personally, I don't think he would have liked what happened with CentOS."
Kurtzer added that "what we are doing with Rocky Linux is really where he would have liked to see the project and open source going. When we named it Rocky Linux, it was a hat tip to him for everything he has done, not only in open source and high-performance computing (HPC), but also with the CentOS project.
"One of the last e-mails that he wrote to the e-mail list was that he was 99 per cent done development of CentOS. It was pretty much ready to go when he passed, but he never saw it released."
The key for an open-source initiative to grow and flourish, said Kurtzer, lies with registering it as a non-profit organization, which was the case with The cAos Foundation. He has done the same with Rocky Linux.
It's official name is the Rocky Enterprise Software Foundation, "backed by an advisory board of trusted individuals and team leads from the Rocky Linux community."
Sounds Good. (Score:3)
But I am still going to let it ripen for a while longer before I commit to a Red Hat based distro.
I still taste the bitterness of CentOS.
Re: (Score:3)
But I am still going to let it ripen for a while longer before I commit to a Red Hat based distro.
I still taste the bitterness of CentOS.
Rocky Linux is never going to get the mindshare penetration that CentOS had. As far as users were concerned, CentOS was free as in beer Red Hat. Rocky Linux is just another Johnny-Come-Lately spinoff distro.
I think the solution to this is what you're indicating: avoidance of Red Hat related distros. Unfortunately, as long as Red Hat remains the corporate standard for Linux, that's going to be a tough job to pull off.
Re:Sounds Good. (Score:5, Insightful)
>"Rocky Linux is never going to get the mindshare penetration that CentOS had."
Hard to say. RedHat screwed the community with what they did with CentOS, and that did sour a LOT of people to anything RHEL based. Many have jumped ship permanently. However, AlmaLinux and RockyLinux are the same as RHEL. And that has a LOT of mindshare behind it. A whole lot.
>"As far as users were concerned, CentOS was free as in beer Red Hat. Rocky Linux is just another Johnny-Come-Lately spinoff distro."
CentOS was a binary-compatible recompile of RHEL (and now isn't).
AlmaLinux is a binary-compatible recompile of RHEL.
RockyLinux is a binary-compatible recompile of RHEL.
The are not any more or less "spinoff distros" than what CentOS was (before it was ruined). They are both open source and "free as in beer" and both have community and commercial support available.
Re: (Score:2)
Re: (Score:2)
>"The difference between Rocky and Alma is that Alma repackages new major RHEL releases several weeks faster as evidenced by v9, so I'd go with them."
What you are saying is true. Cloudlinux already had many years of experience with creating their own RHEL distro (like 12 years). AND they formed the AlmaLinux Foundation before Rocky AND put something out several months earlier. If Rocky gets more popular, it might catch up. Hard to say. Rocky seems like a more "pure" project, which seems attractive.
Re: (Score:1)
in the end, the one that releases sooner, has the resources, and sticks around will likely be the better choice
it is good that there is more than one
Pick one.
Judging a distribution by how fast they release is a poor metric in my opinion. Everyone seems to forget how long it took CentOS to get releases out the door. https://en.wikipedia.org/wiki/... [wikipedia.org] - Take a look at the days for all releases and you'll see what I mean. This isn't unusual. Alma has cloudlinux people basically doing all the work and they already had the system and know how in place. So of course they were ready quicker than others. That's unfortunately how it works. I'm not surprised ei
Re: (Score:2, Informative)
There have been other community defiant shifts in Red Hat that have been poisoning the well,\
* systemd
* Turning CentOS into the beta test platform of RHEL, with un announced and backwards compatibility breaking changes at unpredictable
* The python2/python3 upgrade debacle.
* The python disasters with EPEL
* The multiple and entirely pointless "channels"
* Repeating the Red Hat 9 disaster from nearly 20 years ago with "RHEL 8 Stream" releases, and pretending they don't do point releases. The point releases exis
Re: (Score:2)
>"Cleaning up this mess is going to take work, and I've got clients trying to abandon RHEL every week now."
Neither Alma nor Rocky can clean up any of those (valid) messes, except your second point and sixth, because they are just recompiles of RHEL. Neither project is going to change what goes into their distro or how it works, otherwise, it would not be binary equivalent to an RHEL system.
There are valid reasons for wanting to be based on RHEL (mindshare, support, stability, longevity, experience), but
Re: (Score:2)
We had to endure the pain of porting to Ubuntu because at the time we were forced to decide, neither Rocky nor Alma had shipped yet, and most of the other internal products we work with were already running Ubuntu. Also, one COTS product we use was only ever officially supported on RHEL and Ubuntu so technically we were on our own with CentOS.
The port for us was not that bad; the biggest challenge was removing the muscle memory for typing /etc/sysconfig/network-scripts from the firmware in our fingers. :)
Re: (Score:3)
Rocky Linus is every bit as much a free RHEL as CentOS ever was. Same goes for Alma.
The real question is does anyone really want RedHat at all anymore? It was great way back in the old days when it was just RedHat Linux. It was still OK when it was Fedora Core. But they never got their dependency problems disentangled so you still can't smoothly upgrade between major versions (unlike anything Debian based).
Added to the dependency issues, since going 'Enterprise' They have shown a strong tendency to avoid op
Re: (Score:3)
Yep, I got burned by Redhat's "best effort" on lifecycle twice.
Ported all our stuff to Debian and haven't ever looked back. Almost everything is much easier and there is so much more package support.
Apt pinning is harder than yum/dnf priorities but that's it.
I used to really be hard on Debian (for fair reasons) but kept test machines and Stretch really landed on both feet with Buster cleaning up all the loose ends, surpassing the utility of RHEL.
Re: (Score:2)
You're right... Rocky Linux will likely never be as popular as CentOS. One big reason is that Amazon has its own RHEL clone called Amazon Linux, and it's the default option when you spin up a new EC2 or Workspaces instance on AWS.
Don't underestimate the influence that the biggest cloud hosting provider has on the Linux landscape.
Re: (Score:2)
Unfortunately, Amazon has made some pretty large divergences with Amazon Linux 2. Using python 3.7 rather than python 3.6 profoundly broke a lot of EPEL compatibility. They're also being peculiar with their next Linux release, referring to it as "Amazon Linux 2022" for no sensible reason, since it does not look like it will be released before 2023 at this rate.
https://docs.aws.amazon.com/li... [amazon.com]
It's unclear which RHEL this OS is based on. Despite its claims, it's not based on a co
Re: (Score:2)
Rocky Linux is never going to get the mindshare penetration that CentOS had.
It's just a rebuild, though, so it doesn't really make a squat of difference how many other users there are. If upstream has a lot of users it should be fine, as long as it has some minimum number of build testers, which is a low bar.
Re: (Score:2)
avoidance of Red Hat related distros. Unfortunately, as long as Red Hat remains the corporate standard for Linux, that's going to be a tough job to pull off.
From a certain point of view, that's probably not possible at all. Based on source code provenance reviews, Red Hat is the largest organization contributing development to many of the components in the GNU/Linux stack, all the way from low-level stuff like the kernel, libc, and the compilers, up to user-facing stuff like GNOME. And that's pretty fundamentally why Red Hat is the corporate standard, because those customers want support from the people actually involved in the development of the features tha
Re: (Score:2)
All that past activity did was make me to commit to 100% Debian distros, if not Debian itself. And I saw it coming, especially after the IBM purchase.
Re: (Score:2)
Re: (Score:2)
But I am still going to let it ripen for a while longer before I commit...
I still taste the bitterness of CentOS.
That's what I said, too, but it appears to be ripening nicely IMO, and I'm ready to consider using it going forwards.
Re: (Score:2)
I decided to spice things up and install OpenSUSE. Might have to spin up Rocky and Alma in VM's and see how they act. It's like having a set of LEGO Operating systems. I really love the variety.
AlmaLinux as well (Score:4, Informative)
We went to Alma simply because it was faster out of the gate and seems to be doing a better job of keeping up with Red Hat proper. But with both Rocky and Alma - while both have corporations behind them (yup, Rocky does too), the projects have tried to codify in their governance models that "what happened with CentOS sucked and we're not going to let that happen here".
AlmaLinux (Score:5, Informative)
>'What Happened with CentOS Will Not Happen with Rocky Linux"
Nor with AlmaLinux (by The AlmaLinux OS Foundation), which came out before it, and has perhaps larger support behind it (CloudLinux Inc). We have been using it in production since early this year.
https://almalinux.org/ [almalinux.org]
https://en.wikipedia.org/wiki/... [wikipedia.org]
It is great that there are two strong replacements for CentOS now. The ruin of CentOS really screwed up our plans, but we are back on track now.
Re: (Score:3)
Are there any discussions to merge the AlmaLinux and Rocky Linux distributions?
Re: (Score:3)
There are both costs and benefits to having separate distributions. What would be, best, perhaps, is if they had a commitment to merge their code bases periodically, and kept separate organizations. Perhaps one could try harder to track Red Hat, and the other to harder ensure a stable system with backports for compatibility.
Re: (Score:1)
Re: (Score:1)
I wouldn't be so sure that it couldn't happen to Alma. They are a 501(c)(6) - non-profit **for corporate benefit**. Also, look at who and what it takes to be come a member with vote control on their board. It reads super sketch to me.
Also, https://twitter.com/mattdm/sta... [twitter.com]
Re: (Score:1)
you're not wrong. corporations can buy platinum status which gives them fifty votes on their board per their by-laws. contributors only get one vote. there are multiple tiers of buying votes. it seems to me alma is already on their way to having corporate takeover like centos. the by-laws are not community focused imo. we'll see how well rocky sticks to their principles.
https://almalinux.org/p/founda... [almalinux.org]
Re: (Score:2)
>'What Happened with CentOS Will Not Happen with Rocky Linux"
Nor with AlmaLinux (by The AlmaLinux OS Foundation), which came out before it, and has perhaps larger support behind it (CloudLinux Inc). We have been using it in production since early this year.
https://almalinux.org/ [almalinux.org]
https://en.wikipedia.org/wiki/... [wikipedia.org]
It is great that there are two strong replacements for CentOS now. The ruin of CentOS really screwed up our plans, but we are back on track now.
There's Oracle Linux too. Laugh all you guys want, but it's a well supported direct RHEL derivative with a solid track record. I used to tell naysayers, hey if they pull any tricks, just switch to CentOS, but be pragmatic, Oracle Linux is better supported. Security errata isn't paywalled, they release faster. Well, now it has outlived CentOS. Its history goes back to Oracle wanting to provide an Intel unix platform for Oracle software as a cheap Solaris/Sparc alternative, back when they broke up with S
Re:AlmaLinux (Score:5, Informative)
>"There's Oracle Linux too. Laugh all you guys want, but it's a well supported direct RHEL derivative with a solid track record."
That is true, but Oracle really is evil and had been so for a loooong time. Look what they did to Java, MySQL, OpenOffice, Ksplice, Solaris and VirtualBox for just a few examples of their open-source holdings. I would NEVER use Oracle Linux.
Their latest stunt is to make the VirtualBox "extension pack" addon a commercial license for non-personal use where they won't sell anything less than *100* seats at $50 each. So you are a small business and need 1 or 4 or 10? Tough. And each is LOCKED TO A SINGLE NAMED USER AND A SINGLE VERSION of VirtualBox.
https://shop.oracle.com/apex/f... [oracle.com],
He is going for RH's neck. (Score:2)
RH should watch this guy because this Rocky Linux Project has the capability to be the perfect anti RedHat Enterprise Linux (RHEL) Distro, with effectively the same business model but no profit motivation so they will be beating them every time depending on how fast they can ramp up.
Imagine what would happen if just one of RHEL's good quarters of profit from the glory years of old had been injected into a single open source project fully without ANY profit seeking spend. Just only dev salaries working full
Re: (Score:2)
Huh? Rocky CAN'T out-innovate RHEL; it's a binary-compatible clone of RHEL.
Re: (Score:2)
You seem to be (hopefully by mistake) misunderstanding my post.
Are you seriously not seeing the opportunity here?
Re: (Score:2)
You don't seem to have a fucking clue. First, Red Hat does not sell Linux. They sell SUPPORT of Linux to customers who expect guarantees of things working. They do this by building a distribution and then applied fixes only to it for 5 years or so. That takes actual work. Neither CentOS nor Rocky either sell support nor do the work of back-porting fixes, etc into a stable release. All they do is take the work Red Hat has done (because it is GPL), replace any copyrighted artwork and trademarks with thei
Re: (Score:1)
Opportunity for what exactly?
There are dozens, if not hundreds, of Linux distros already. What special thing does what you imagine have? What will it do?
Just fork Fedora (Score:2)
Huh? Rocky CAN'T out-innovate RHEL; it's a binary-compatible clone of RHEL.
So true. Anyone that wants to out-innovate RHEL should fork Fedora, or maybe take over it so that it produces RHEL quality releases.
Re: (Score:2)
I would argue that the profit motive is what drove Red Hat to produce the best Linux distro available. It was IBM and Red Hat (before they merged) that did the most to make Linux what it is today. Other corporations were only supportive of Linux because those two IBM legitimized it and Red Hat made it commercially viable.
So you want a community-centric Linux? Use Arch or something Arch-based. But the corporate world won't touch it. They want support, which means Red Hat, Canonical, Cloud Linux, SuSe or some
Re: (Score:2)
RH should watch this guy because this Rocky Linux Project has the capability to be the perfect anti RedHat Enterprise Linux (RHEL) Distro, with effectively the same business model but no profit motivation
The RHEL business model is to sell commercial support, and you can't sell that without a profit motive.
I prefer Debian to IBM, thanks very much. (Score:4, Interesting)
It was too easy for Red Hat do do what it did and every distro based on Red Hat is based on what is now an IBM product.
Re: (Score:1)
Yes I'm also using Debian in all my servers, however when RedHat says jump almost all the other distros say how high, most of the time including Debian. So it's getting harder to run away from RedHat(ism) every day.
Re: (Score:2)
After commenters, see why never tried Linux (Score:2)
Re: (Score:2)
Hey bozo, please tell me which version of currently supported Windows you've decided to use:
Windows 10 Home
Windows 10 Pro
Windows 10 Education
Windows 10 Enterprise
Windows 10 Pro for Workstations
Windows 10 Pro Education
Windows 10 S
Windows 10 Enterprise LTSC
Windows 11 Home
Windows 11 Pro
Windows 11 Pro for Workstations
Windows 11 Pro Education
Windows 11 Education
Windows 11 Enterprise
Windows 11 SE
Thanks
Re: (Score:2)
Re: (Score:2)
Good answer :)
OSX then? They really don't have multiple different versions of OSX.
AlmaLinux is where we ended up (Score:1)
AlmaLinux has a lot of really good features that go beyond CentOS/Rocky options. We ended up going this direction for the in-place migration strategy that Alma supports.