Red Hat to move focus away from CentOS in favour of Stream; CentOS team discuss implications with Wikinews
Monday, December 14, 2020
Image: CentOS.
On Tuesday, US-based software company Red Hat announced their plans to shift their focus away from CentOS in favour of CentOS Stream.
Image: Red Hat.
In the blog post Chris Wright, the Chief Technological Officer of Red Hat said Red Hat had informed the CentOS governing board that Red Hat was "shifting our investment fully from CentOS Linux to CentOS Stream". At least five of the nine CentOS Governing Board listed on the project are Red Hat employees.
Started in 2004, CentOS has been a free-of-cost free/libre open source software which provided binary-code compatibility with Red Hat Enterprise Linux (RHEL) — Red Hat's GNU General Public Licensed paid operating system. CentOS was absorbed into Red Hat in 2014, with Red Hat gaining the trademark rights of "CentOS".
Red Hat also sponsors the Fedora operating system. Red Hat even gives the software engineering interns laptops with Fedora on it. Till now, software development took place on Fedora, which was later adopted in RHEL, which the Red Hat maintained and provided support for, for those customers who had RHEL subscription. CentOS would then follow RHEL to provide the same features free of cost, but without the support.
Stream was announced in September 2019, just two months after Red Hat was acquired by IBM. CentOS Stream's development cycle had new features added to it before the features became a part of RHEL. Stream receives more frequent updates, however, it does not follow RHEL's release cycle.
With CentOS Stream, developments from the community and the Red Hat emplyees would take place beforehand on both Fedora, and Stream as a rolling release, before those features are absorbed into RHEL. CentOS followed the release cycle of RHEL and therefore it was a stable distribution. Features available in CentOS were tried and tested by Fedora, and then RHEL maintainers. Stream, on the other hand, has features added to it before those features become a part of RHEL. That implies Stream would be ahead of RHEL's development, containing new features which are not yet tried and tested by RHEL developers, and not be binary code-compatible with RHEL.
Writing "The future of CentOS Linux is CentOS Stream", Wright further wrote in the announcement, CentOS Stream "provides a platform for rapid innovation at the community level but with a stable enough base to understand production dynamics." Wright also said, "CentOS Stream isn't a replacement for CentOS Linux; rather, it's a natural, inevitable next step intended to fulfill the project's goal of furthering enterprise Linux innovation."
Since the announcement was made, many people expressed their anger on Internet Relay Chat (IRC), Reddit and CentOS project's mailing list. CentOS 8's End of Life (EOL) has been moved up from May 2029 to December 31, 2021, while CentOS 7 is expected to receive maintenance updates through June 2024, outliving CentOS 8.
Image: Gregory Kurtzer.
Gregory Kurtzer, who had started the CentOS project, announced a new operating system, Rocky Linux. Rocky Linux describes itself as "a community enterprise Operating System designed to be 100% bug-for-bug compatible with Enterprise Linux". The project said "Rocky Linux aims to function as a downstream build as CentOS had done previously, building releases after they have been added to the upstream vendor, not before." The project maintainers have not decided a date for when the operating system will be released.
Kurtzer explained the reason to call the OS as Rocky Linux: "Thinking back to early CentOS days... My cofounder was Rocky McGaugh. He is no longer with us, so as a H/T [hat tip] to him, who never got to see the success that CentOS came to be, I introduce to you...Rocky Linux".
Image: Red Hat.
Image: Pablo Greco.
Wikinews reached out to the members of the CentOS project, Pablo Greco and Rich Bowen, to discuss this move, its implications as well as the future of CentOS. Greco is a maintainer of armhfp (32-bit processors) and is a part of the CentOS' quality assurance team. Bowen — who unlike Greco, works for Red Hat — is the community manager of the CentOS project.
This story has updates
See Gregory Kurtzer discusses plans for Rocky Linux with Wikinews as Red Hat announces moving focus away from CentOS, December 18, 2020
Interview with the CentOS team
Image: Evan-Amos.
Can you tell me about your current role in the CentOS development, and how did you get involved with the project?
((Pablo Greco
)) I'm the armhfp maintainer, basically helping out fixing things that are not part of the Red Hat supported architectures. I do help out with other things like secondary kernels, and creating images for armhfp and aarch64. I got involved not too long ago when a project that I was working on needed to use ARMv7 hardware, and although CentOS had a version, it wasn't complete.
((Rich Bowen)) I'm Rich Bowen and I'm the Community Manager for the CentOS project. That means that it's my job to help people get involved in the project, manage the social media and newsletters (the "voice" of the project), and generally work with the community. I've been at Red Hat for almost 8 years, and during that time have worked in this capacity with a couple of other projects, too. I've been dedicated to the CentOS Project for going on 3 years now. And I've been doing open source stuff for about 25 years now.
I should also note that I'm not alone — either on this team, or in responding to these questions. I have a team of colleagues who help me out with CentOS through the Open Source Program Office (OSPO) at Red Hat, and across the many other departments that are part of the Fedora/Red Hat Enterprise Linux (RHEL)/CentOS projects at Red Hat.
There's also a number of colleagues looking over my shoulder as I write this response, to ensure that I get the story straight. As a community manager, I'm not the technical expert on all the things, and so I need that support to get the details right in some of your questions.
((WN)) What is the relationship between CentOS and RHEL?
((Pablo Greco)) CentOS started as an external project, that was acquired by Red Hat 7 or 8 years ago. But I don't think that it counts as an answer. I think you'll get a better answer from people from Red Hat.
((Rich Bowen)) In many ways, it's the same as it's always been, but with a different end result and project structure. This is a community-focused open source project that delivers a strong Linux distribution foundation where non-operating-system focused ("layered") open source projects can develop, test, and innovate. For example, Ceph and RDO.
The main shift with this evolution of a focus on CentOS Stream is how it is developed in the open, in the community space, allowing contributors to get involved in the creation and shaping of the future. This also means RHEL users, customers, partners, independent hardware/software vendors, and anyone else can participate in the future of RHEL in true open source fashion.
In the past, the relationship between Fedora, RHEL, and CentOS was effectively, "Fedora gets snapshot every so many years, then RHEL engineers develop RHEL behind closed doors until release day, then the RHEL source code is released, then CentOS would take the RHEL source and rebuild." While that is an oversimplification, it is essentially accurate. Now the new structure is, "Fedora gets snapshot every so often and becomes the basis for the next CentOS Stream release, then development happens in the open with community contribution involvement and influence, and then RHEL rebases frequently on CentOS Stream."
This moment of integration is called "the development phase" in the RHEL engineering process. This is being misinterpreted as a "beta". Rather, this is an aspect of a distribution that is continuously delivered.
This is not software that is untested and not considered ready for inclusion in RHEL. It is already planned for the next RHEL release, has passed QA [quality assurance] and CI [Continuous Integration] checks, and is considered stable. In the past, this software went right into RHEL without external visibility. We don't schedule software for inclusion into RHEL unless it has been used/tested/integrated at other levels (such as Fedora), and is considered stable enough for inclusion. This same philosophy will still be at play with CentOS Stream.
One thing I would like to highlight here is this is the first time in the history of RHEL that development happens in the open and allows for collaboration and contribution from the broader open source community. That is a large goal of CentOS Stream and we're making the move to focus on that entirely.
Image: Fedora Project.
For an example of where this evolution has been discussed in the past, this presentation from Fedora's Flock 2018 "The Penrose Triangle" shows how this problem has been discussed and worked through for a number of years now.
Image: Paul Rand.
((WN)) How has CentOS evolved after Red Hat acquired it? How has it evolved since IBM acquired Red Hat last year?
((Rich Bowen)) The biggest change when CentOS came to Red Hat was that individuals who were working on CentOS in their weekends and evenings now had the ability to devote more time to it.
In addition, the project for the first time was able to put significant resources on helping make the platform useful for open source software development of other projects in Red Hat's ecosystem. This included the expansion of the SIG program [Special Interest Groups] to include layered projects like Gluster, and OpenStack.
IBM is not involved in the CentOS community or Red Hat's work in it. An unfortunate misconception about this announcement has been conspiracy theories about IBM's involvement in this decision. Red Hat operates independently from IBM, and IBM had no influence or involvement in the decision.
Image: Wondigoma.
((WN)) What are the similarities between CentOS, Fedora and RHEL?
((Pablo Greco)) I'm gonna explain in a really loose way, ok? Fedora is where everything starts, all the fast moving packages. Some maintained by Red Hat employees, some by contributors. Every few years, Red Hat gets sort of a snapshot of Fedora, and starts stabilizing and improving it, and that becomes RHEL. CentOS originally was a rebuild of that.
About 1 year ago, Stream was introduced, as a way to make the work around RHEL more public, allowing the users to contribute. Making it sort of a continuous preview of what's to come for RHEL. That doesn't imply anything about it being of lower quality by definition, or some continuous beta like some people call it. Many of the packages that are built for stream, reach the next version of RHEL unmodified. So even if there is a higher level of testing in Stream, it is far from a traditional beta.
((Rich Bowen)) They are each Linux distribution projects, which work on a common code base at different points in its life. It's born in an upstream open source project, and travels through each of the projects as it matures.
From a community perspective, the projects are each run as open, collaborative communities, but with different audiences. By the time we get to RHEL, it's done inside Red Hat, but that team also operates with the principles of open source.
The exciting thing about CentOS Stream is that it makes RHEL development more open, both in terms of decision making, and in who can participate. We're believers in the Four Opens, as the OpenStack community talks about them here — https://governance.openstack.org/tc/reference/opens.html
Our roots are in open source, but that's not enough without the other three.
((WN)) What are the similarities and differences between CentOS and CentOS Stream? Or do they just share the name and are completely different?
((Pablo Greco)) For me, Stream is not CentOS, Stream is the next RHEL, the rest is just branding.
((Rich Bowen)) The name "CentOS" is frequently misused, and has been for a long time, so here are the basics: CentOS is the project. CentOS Linux is a rebuild of Red Hat Enterprise Linux (RHEL). CentOS Stream is an upstream distribution from RHEL, which reflects what the RHEL team intend to put into the next minor release of RHEL. The CentOS Project also contains other things, such as the SIGs (Special Interest Groups), the Promo/Events team, the CI/CD and build system (CBS) project, and so on.
((WN)) Do you have any stats for the userbase of CentOS vs Stream?
((Pablo Greco)) I don't, but Stream is rather new so I don't expect the number to be that big. I've been using Stream for most of my things ever since it came out, and I love the idea for it. The only problem is that Stream is being used as an excuse to kill CentOS, and that's the part I hate.
((Rich Bowen)) We don't. For reasons of privacy, we have never kept server logs around CentOS Linux, and other CentOS Project "downloads." And these files are served by a large mirror network distributed around the world, much of which we don't have any actual access to.
I put the word downloads in quotes, because installing a Linux distribution isn't exactly just downloading something. It's composed of thousands of packages which are (can be) installed independently, and figuring out which of those packages are going to CentOS Stream machines, vs CentOS Linux machines (vs Oracle Linux machines, and so on) and which ones are being mirrored for further use by additional users behind a firewall, is a lot of guesswork.
We have some general ideas, but have always been reluctant to try to put actual numbers on it, because of these reasons and others. Linux user surveys [...] perhaps have a better handle on this than we do, since we have simply never gathered these stats.
((WN)) Could you briefly explain what Red Hat announced and what it means for the future of CentOS?
((Pablo Greco)) They announced that the CentOS as we all know it, doesn't exist anymore, there's only Stream from now on. This applies to all releases starting with [CentOS] 8, [CentOS] 7 seems to be unaffected by this. In my view, CentOS is dead, there is no CentOS anymore, because as I said, I view Stream as a RHEL thing and not a CentOS thing.
((WN)) Could you briefly explain Chris Wright's announcement, and what it means for the future of CentOS?
((Rich Bowen)) The future of CentOS is stronger than ever, as it is now literally in the critical path of RHEL development. The CentOS Project as a community continues, and in fact the community grows more facets as CentOS Stream provides more ways where the community can engage with Red Hat and each other. There will be some growing pains as users of CentOS Linux find out and help shape whether CentOS Stream meets their needs. Communities are built in overcoming challenges together, and there are many challenges ahead in making CentOS Stream right for members of the CentOS community.
A major aspect of this is that in the past, CentOS Linux users could not shape the future of the distro within the project. From now on, not only developers but end users of all types will be able to shape the future of the distro. If CentOS Stream does not meet a user's need today, now they can advocate and contribute toward it meeting their needs as it evolves.
((WN)) Did Red Hat speak with the CentOS core developer team about this plan?
((Pablo Greco)) I don't have a contractual relation with Red Hat, just as many of the people that helped in the CentOS-QA team. I assume that those who are knew about this, but the rest of us didn't. But this is all my assumption, don't have any evidence one way or the other.
((WN)) Did Red Hat speak with the CentOS core developer team about this plan? Were there any oppositions within Red Hat about this decision? Was it put to vote?
((Rich Bowen)) Red Hat discussed this with the CentOS Governing Board, which is responsible for the oversight of the CentOS Project. There was a lot of back-and-forth on that proposal. The CentOS Governing Board voted and reached consensus.
Red Hat's engineering resource focus has never been decided by voting. Some of the core developers are on the Governing Board but not everyone in our community was involved in the decision. As I'll touch on later in this interview, that's one of the reasons that some of the community have taken this change so hard.
Is every Red Hatter happy with the announcement? Of course not. A big part of Red Hat's culture is the ability to voice an opinion and ask hard questions. And Red Hatters have definitely expressed their thoughts about this.
((WN)) Do you know why this sudden move from Red Hat?
((Pablo Greco)) No idea, I can assume it was a business decision, but no information whatsoever.
((Rich Bowen)) CentOS Stream was introduced in September 2019, so while the news may have come as a surprise to many, Red Hat has also had that time to get feedback on CentOS Stream and see momentum that reinforced the role it believed CentOS Stream could fit in the RHEL ecosystem. I've seen Red Hat's VP of Linux Engineering explain the timing internally — once Red Hat took the decision to the CentOS Governing Board and the Board had voted, they all thought it was important and most transparent for the community leadership to communicate that decision as soon as possible.
((WN)) Why has the End of Life [EOL] been shortened?
((Pablo Greco)) Good question, I don't have an answer for it. I guess someone from Red Hat should give you that. I'm not happy about it to say the least.
((WN)) Why has the EOL been shortened? Is this unusual?
((Rich Bowen)) First, it's important to understand that the dates that the CentOS project lists as EOL (End Of Life) dates are, and always have been, dependent on Red Hat. That is, we can say that CentOS Linux 7 will be EOL on a particular date, but if the dates around RHEL 7 shift, hypothetically speaking, then so will those around CentOS Linux 7. That's what happened with CentOS Linux 8. This decision was made by Red Hat and the CentOS Board.
((WN)) What are some of the reasons people consider CentOS for their daily driver or for their servers?
((Pablo Greco)) It used to be rock solid distribution, with almost everything that RHEL has, but without all the hassle or the costs of a RHEL license, and obviously without the main benefit of RHEL which is the support. CentOS 7 still fits that description.
((Rich Bowen)) Besides the obvious cost benefits of running CentOS Linux, many users find that the friction of running CentOS Linux is much lower than other options. There's no web form, registration, or subscription management. And while these frictions are things that the RHEL subscription team is actively working on, this remains part of our users' mindset.
((WN)) How different are CentOS 7 and CentOS 8?
((Rich Bowen)) Upgrades between major release versions provide a chance to provide the latest versions of packages to our users. A major release will stay on these base package versions for the releases' lifecycle. The "rebasing" of packages allows us to incorporate all of the fixes and enhancements that have occurred between major releases and ensure we maintain the balance of upstream momentum, with the stability that backporting patches and fixes provides.
((Pablo Greco)) In some ways, very different, Modularity may be the biggest, many new versions of things (there was a 5 year gap between 7 and 8). But for the most part, usage is the same as it has been forever, like in things like repos [repositories], installing and removing packages, building packages, etc.
((WN
)) Anyone who would be running CentOS 7 could upgrade to CentOS 8 without any breaking changes?
- ((Pablo Greco)) There are several attempts to do that, none successful as far as I know, I'm really hoping to be wrong.
((WN)) Would it be feasible for the developers to fork CentOS 8 after EOL, to continue developing the OS?
((Rich Bowen)) We've actually covered this in the FAQ that we published on Tuesday. The short answer is yes, and we fully expect that to happen. Indeed, the Rocky Linux project has already indicated their intention to do just that.
There's also ongoing discussion on the CentOS Project mailing lists of how this could be done within the community itself.
((Pablo Greco)) Very, to me the most difficult part of coming up with a new fork is the human factor. The technical side may have its complications (many to mention), but at the end of the day, they are solvable.
- ((WN)) The human factor?
- ((Pablo Greco)) People interaction, who does what, responsibilities, jealousy, etc. Dealing with people is a lot harder than dealing with computers.
- ((WN)) Are there any discussions taking place in the community to fork CentOS?
- ((Pablo Greco)) As far as I know, Rockylinux is the main one, it has gotten some traction, hopefully something good will come out of it. Then there is CloudLinux, which announced that during Q1 2021 they'd be releasing their rebuild for free.
- ((WN)) Is CentOS Stream not as rock solid distribution as CentOS?
- ((Pablo Greco)) No, and that's not a bad thing. just for different use.
- ((WN)) What are some of the use cases for which CentOS is ideal?
- ((Pablo Greco)) Personally, I use CentOS for everything, [CentOS] 7 for some servers, [CentOS] 8 for others, Stream for development and desktops. But that's just me, many people could use it in different ways.
- ((Rich Bowen)) CentOS Stream is ideal in many (even most) of the same situations that CentOS Linux is good for.
- Look, with my Red Hat on, I want everyone to use RHEL, because it puts food on my table. But I am also passionate about the CentOS platform as a solution to everyday computing needs, from a desktop machine to your server farm. And Red Hat wants people to adopt CentOS Stream widely, so that it can actually realize its benefit to the RHEL development program.
- To that end, we'll be trying to make CentOS Stream as widely useful as possible, and as stable as possible, so that people actually use it. Because if they don't use it, this was all for nothing.
((WN)) Will Red Hat employees be barred from contributing to any community-driven development of forks of CentOS?
((Rich Bowen)) A large part of Red Hat's culture is ingrained in the idea of open source and community. Red Hat will never prevent its employees from pursuing upstream development and open source projects.
I think section 6 [of Red Hat's Open source participation guidelines] "Contributing to upstream projects not maintained by Red Hat" applies here.
((WN)) What are the implications of this move for the userbase?
((Pablo Greco)) Where to start? There are too many cases in the userbase to mention, but taking some examples. If you have a fast deploying platform (like the Facebook example given in the press release), and you take advantage of the benefit of Stream even with the shortened EOL, this may not even affect you. But in my opinion, this is not the majority. Most users wanted CentOS for its premise of being bug-for-bug compatible with RHEL, and since that option doesn't exist anymore, until a suitable replacement appears, things are going to be complicated. I guess I could extend this by a lot, but that's the gist.
Depending on your use case, you may be unaffected, or have a really complicated year ahead of you. I think very few people will be unaffected.
- ((WN)) "being bug-for-bug compatible with RHEL" -- Is it fair to call CentOS as RHEL, but without the support?
- ((Pablo Greco)) Fair, but not entirely accurate. RHEL customers have access to tools that may not be public, and therefore, not available for CentOS. I mean, there are other benefits of being a RHEL customer, aside from support. Not all the tools are code, there's access to different containers, knowledge base, etc.
- ((WN)) If I understand correctly, Stream is going to be a RHEL beta, and a rolling release, similar to Arch [Linux] or [Open]SuSE's Tumbleweed. Right?
- ((Pablo Greco)) Not exactly. Let's say it is a minor-release-rolling, that's how I'd describe it. It is not a continuous rolling like Arch. And with respect to beta, there's never code that is known to break, everything that lands in Stream is expected to be shipped as-is in the next RHEL minor release. That expectation may not always be met, but still.
((WN)) Is the success of Stream somehow responsible for this move?
((Pablo Greco)) In my opinion, no. The need for Stream to succeed maybe, but I just don't see Stream succeeding under the new conditions.
- ((WN)) Oh, why do you say so?
- ((Pablo Greco)) Because I've lost all my incentives to collaborate in Stream. The idea was that it would be a preview of what's to come, and that's great. But they just removed the end-result of Stream, which is the next minor release. Stream helps make a better product, but we don't have that product anymore (CentOS Linux), so.
((WN)) Is the success of CentOS Stream somehow responsible for this? In the announcement, Wright said Stream was not the replacement for CentOS. What is the possible solution for those who relied on CentOS?
((Rich Bowen)) CentOS Stream is not a drop-in replacement for all existing use cases. It requires sysadmins out there to evaluate the new information and adjust their workflows. We would do our user base a disservice to claim otherwise. For many use cases it can be used as a replacement. Indeed, for people who wanted the stability of the CentOS Linux base operating system, but were frustrated by the slowness of application updates, CentOS Stream will be a better replacement for CentOS Linux — now those updates will be available sooner, should users choose to update.
We also want people to understand the difference, so they can see the additional options it brings to them. We want people to go into this decision with eyes open, and not expect it to be something that it's not.
((WN)) What can be done for those users for whom CentOS Stream does not meet their requirements?
((Pablo Greco)) Pay for RHEL, move to another distribution, or wait for someone to fill in the gap. If someone has another solution, I'm all ears.
((Rich Bowen)) We hope any users who are in that situation will tell us, so we can figure out how to fix that failing. One of the exciting things about the move to CentOS Stream is it's designed to be an actual contributor community, rather than the legacy CentOS Linux that is almost entirely a consumer community.
CentOS Stream is more than just something you download and install. It's a process and a community — a venue for contributing changes, making suggestions, and reporting bugs. All of that feeds back into making CentOS Stream better, and, in turn, making RHEL better.
That said, Red Hat is working on some "low- or no-cost programs" for users who cannot, for whatever reason, deploy CentOS Stream in their environment. This is discussed in the Red Hat announcement and we anticipate hearing more details around that in the coming weeks.
((WN)) What are some of the best arguments in favour of, and against this move that you are aware of?
((Pablo Greco)) I have nothing positive to say about the announcement, let's just leave it at that. [About against the move,] I can't put it into words you can publish.
((Rich Bowen)) While I empathize with users' response to the decision, I believe that shifting focus to CentOS Stream is the right solution for our community, in the longer term.
CentOS Stream creates a place where the community can actually participate in the development of the OS on which they rely. They can now actually file bugs against it and have those considered by the engineering team. They can submit their own changes to the project and expect that they will actually be considered and accepted into the code.
The arguments against it have been made eloquently elsewhere, and if I can summarize them, they would be timing (dropping the announcement right on the heels of the CentOS 6 Linux EOL) and timing (cutting the maintenance lifetime of CentOS Linux 8 from 10 years to 2).
((WN)) How will this move affect the development of RHEL, and other GNU/Linux?
((Rich Bowen)) RHEL maintainers try to do as much work as possible "upstream first". This means starting in upstream open source projects like the kernel, then participating in Fedora, then bringing those changes in for RHEL. CentOS Stream gives RHEL maintainers a place to collaborate with the community in public on what's coming for the next minor release.
((Pablo Greco)) I don't think this will affect other distros in a negative way, quite the contrary. With respect to RHEL development, depends on how much people contribute to Stream. If people contribute to it, it will make RHEL a better product, because it will have a broader testing in the earlier stages. But that's just the benefit of stream. I mean, this was still true before the announcement.
- ((WN)) Hm, how can it benefit other distros? I mean, how does the development of any distribution benefit other distros?
- ((Pablo Greco)) For me, it is going to benefit other distros because many people, angry at this announcement, will move away from the RHEL ecosystem. And that could mean more resources, but it is all just guesses.
((WN)) Will the maintainers appoint some new maintainers (not necessarily Red Hat employees) to see some development? Can the current maintainers continue to work on CentOS in their spare time?
((Pablo Greco)) As far as I understand, no. CentOS is a trademark of Red Hat, so anything would have to be done on the outside. But I'm not a lawyer, and I think you should get a formal answer to that from someone who represents Red Hat.
((WN)) What does the strong reaction from the CentOS community indicate?
((Pablo Greco)) For me it indicates that Red Hat didn't expect this, or that they did expect it and were prepared to handle it. I don't like any of the options.
((WN)) What is the reason for such a strong reaction from the CentOS community? Had Red Hat considered this while making this move? How are they planning to handle it?
((Rich Bowen)) Open source communities are often very passionate about their projects. The very nature of open source means that each person has equal ownership of the project. So the strong reaction here is understandable. A decision was made outside of their view, which they would not have made themselves. That makes people feel betrayed. We get that.
And, yes, we discussed, in the leadup to this, how it would affect people, and that they would be angry. We tried to minimize that, but we understand community skepticism about that statement. It's a big change, being made quickly, and driven by business needs that the majority of our user community doesn't know or particularly care about.
I've talked in other places about the stages of grief. The community has lost something, and they are grieving that. Part of that process is anger. This is neither unexpected nor strange in any way.
We also understand that this explanation rings hollow to some, and sounds like, "We considered that and then decided to ignore it." That's not the case, but we understand why some people feel that it is.
How are we planning to handle it? By ensuring that CentOS Stream lives up to our promises by doing rather than talking. I'm here, as the community manager, to help people find the solutions that they need, and to help them navigate this period of transition.
((WN)) How many core developers are there, who are not working for Red Hat?
((Pablo Greco)) The CentOS QA team which handles most of the day-to-day things, are about 20 people, most of which are not Red Hat employees, like myself.
- ((WN)) What lies ahead for you (and the non-Red Hat developers)?
- ((Pablo Greco)) I can't speak for the others aside from the general anger. I'm considering my options, don't want to hurry and say something that I'd have to retract. Right now, the working theory is that I'm going to help with CentOS 7, because it is still alive, and in fact, it will outlive CentOS 8. But I don't plan on reaching a conclusion this year.
- ((WN)) Did the non-Red Hat developers collectively send a letter/email to Red Hat about this?
- ((Pablo Greco)) Not that I know of. And not that it would change anything in my opinion, this was planned for a long time. The way I see it, Red Hat doesn't view the non-Red Hat employees as a part of CentOS, so we're just like anybody else.
((WN)) How do you think Fedora now fits in the grand scheme of Red Hat development?
((Pablo Greco)) Fedora is still the upstream for whatever comes to RHEL, and is where most of the big changes happen. So in my opinion, Red Hat needs a stable and vibrant Fedora.
- ((WN)) Isn't Stream achieving the same goal?
- ((Pablo Greco)) Different stages. Take 8 for example. 8-stream is the next minor version of 8. When RHEL takes a snapshot of Fedora in a few months, that will become 9 stream. And Fedora will continue moving. Part of the preparation for the snapshot is being done in fedora as ELN (Enterprise Linux Next).
((WN)) Many users are calling it "The death of CentOS". Do you agree with that? Will the maintainers appoint some new maintainers (not necessarily Red Hat employees) to see some development? Can the current maintainers continue to work on CentOS in their spare time?
((Rich Bowen)) No, obviously I do not agree that this is the death of CentOS. As I said earlier, CentOS is a project with multiple outputs, and the remaining outputs are still important both to CentOS users and to Red Hat. The project continues, it just looks different now, with a different primary focus.
That said, we totally understand community members who feel that we have killed the part of the project that they care about, and that, to them, this means the death of the project. That's not confusing or surprising at all.
In the coming year, we hope to demonstrate that CentOS Stream is every bit as useful, to most of our users, as CentOS Linux was. And for the users for whom it is not as useful, we will strive to understand why, and whether we can address that gap.
The very first sentence of my blog post on Tuesday was "The future of the CentOS Project is CentOS Stream." Clearly that implies that I don't believe that this is the end of the line, or the death of the project. Our job going forward is to demonstrate the value of CentOS Stream, and build the contributor community around it.
((WN)) How can an open source project ensure situations like these do not repeat?
((Pablo Greco)) Wish I knew, maybe not letting for-profit companies dictate the future of non-profit organizations/distributions. Unfortunately, I don't think I'm equipped to solve that :(
((Rich Bowen)) All projects in the open source ecosystem are unique and face unique challenges. We would not presume to try to predict how to anticipate another project's issues.
Sources
- "CentOS Product Specifications" — CentOS, December 12, 2020 (date of access)
- "Governance" — CentOS, December 12, 2020 (date of access)
- Chris Wright. "CentOS Stream: Building an innovative future for enterprise Linux" — Red Hat, December 8, 2020
- Rich Bowen. "CentOS Project shifts focus to CentOS Stream" — CentOS, December 8, 2020
- Abhishek Prakash. "IBM's Red Hat Just Killed CentOS as we Know it: With CentOS Stream, Stability Goes out of the Door" — It's FOSS, December 9, 2020
- Chris Wright. "Transforming the development experience within CentOS" — Red Hat, September 24, 2019
- "IBM Closes Landmark Acquisition of Red Hat for $34 Billion; Defines Open, Hybrid Cloud Future" — Red Hat, July 9, 2019
- "Red Hat and CentOS Join Forces" — Red Hat, January 8, 2014
- "Rocky Linux" — GitHub, December 12, 2020
External links
- "Open source participation guidelines" — Red Hat, December 12, 2020 (date of access)
- "FAQ - CentOS Project shifts focus to CentOS Stream" — CentOS, December 12, 2020 (date of access)
- "The CentOS SIGs" — CentOS, December 12, 2020 (date of access)
- "The Four Opens" — OpenStack, December 12, 2020 (date of access)
- "December 2020 Archives by thread" — CentOS, December 12, 2020 (date of access)
- "Rocky Linux" — Rocky Linux, December 12, 2020 (date of access)
- Brendan Conoboy. "How RHEL is Made" — CentOS, December 11, 2020
- "FAQ: CentOS Stream Updates" — Red Hat, December 8, 2020
- Stef Walter. "CentOS Stream is Continuous Delivery" — CentOS, December 11, 2020
- Thomas Oulevey. "Minutes for CentOS Board of Directors for 2020-11-11" — CentOS, December 11, 2020
