I have been forced to be a bit harsh right now based on how much I love RHEL and other Red Hat products. It is because I want RHEL [and/or its clones] to be accessible to everyone.
Excuse my tone this time because my heart doesn't want RHEL to be inaccessible to hobbyists. The free RHEL developer license is just whataboutism and dilutes the enthusiasm felt by the broader hobbyists' and tinkerers' community compared to when true RHEL clones existed.
I, with my use case, am happy using RHEL with the free subscription. In fact, that's what I'm using right now. This article is about what happens to RHEL if this is maintained.
A Quick Recap of the Events That Led to This
I'm sure you all know by now why Red Hat has been in the news recently:
But, if you have no idea what's going on, lemme give you pointers to a quick timeline that should help you understand:
- Red Hat has an excellent Linux distribution called Red Hat Enterprise Linux with a support cycle of 10 years.
- Like any other Linux distribution, the source code used to build RHEL (or a clone of RHEL) was available publicly.
- CentOS came into existence by taking the aforementioned source, removing trademarks, and creating an almost 1-to-1 copy of RHEL. (I say almost because RHEL trademarks need to be removed.)
- The enthusiast community--that didn't bother paying fees just to try out RHEL for the first time--can now use CentOS to see if they like it.
- A subset of people from that group then inform their higher management about RHEL and how good it is based on their initial experience with CentOS and most likely switch to cater to either supporting their products on RHEL or using RHEL for their deployments.
- The enthusiast community grows because CentOS is free. People like Jeff Geerling use CentOS to teach up-and-coming enthusiasts interested in cosplaying as a sysadmin about awesome stuff like Ansible with his open source repositories.
- CentOS's binaries were released a few weeks after RHEL's updates. So Red Hat acquired CentOS to ensure that CentOS didn't lag behind RHEL's updates.
- A few years later the CentOS Stream fiasco happened, killing CentOS.
- Rocky Linux and Alma Linux come up to fill the void left by CentOS. They use the RHEL's source which is publicly available.
- Red Hat starts restricting the source code to its customers and anyone else with a free developer subscription.
I specifically didn't mention IBM buying Red Hat because, if someone from Red Hat has the guts to publicly say, "Simply rebuilding code, without adding value or changing it in any way, represents a real threat to open source companies everywhere." in an official capacity, they can also be trusted to be blunt about IBM's involvement in these decisions. But almost every Red Hat employee has publicly denied this accusation. Therefore, I don't believe IBM shot themselves in their foot either. But you are free to believe what your heart says.
And no, I'm not being sarcastic when I say IBM is probably not involved in this decision. They may be, but I don't think so.
Dear Red Hat, What Did You Just Do?
You offer a product to the community for free (CentOS). Then you switch the support window of that free product mid-cycle (CentOS) and use its "alternative" (CentOS Stream) as a testing ground for your shiny, enterprise product (RHEL).
The leeches then take the source code of your shiny product and create a successor to CentOS (Rocky Linux and Alma Linux). You don't like that so you "soft-paywall" them.
Now while you are well within your rights to do so (because you earn money from providing support for RHEL and not from RHEL itself), I'll explain why this is a bad move for yourself.
Question: Why is RHEL so popular in enterprise environments?
Question: What did online tutorials use to teach about RHEL?
Question: What do the followers of said online tutorials use to learn about RHEL?
Question: When said student becomes a teacher, what will they recommend to someone who asks how to get started as a Linux sysadmin?
Question: If someone is willing to pay for RHEL's license but is hesitant to use it because of non-public public software repository, what will they use?
In a nutshell, if Red Hat continues to cause trouble for RHEL clones, here's what you can expect:
- Many customers and professionals helping with enterprise deployments will consider abandoning RHEL and not supporting it.
- New users will start considering Ubuntu, Debian, openSUSE or something else which is there to stay.
- Universities/IT Courses will also switch to replacements like Ubuntu or openSUSE instead of dabbling between RHEL clones, CentOS Stream, and Fedora.
Moreover, small businesses and universities cannot use the free Red Hat Developer Subscription as per their FAQ:
The no-cost self-supported Red Hat Developer Subscription for Individuals is designed for individuals and personal accounts.
Unless Red Hat is friendly towards RHEL clones, I don't see any newcomers in this ecosystem anymore. And, this is just depressing, as RHEL is a fantastic product. Yes, it's not as bleeding edge as even Fedora, but man it is still fun to use!
I will let this quote from Brian Stevens be here...
It is core to our beliefs that when people who share goals or problems are free to connect and work together, their pooled innovations can change the world. We believe the open source development process produces better code, and a community of users creates an audience that makes code impactful.
RHEL is an enterprise distribution so Red Hat will almost never work at providing support for running RHEL on Raspberry Pi. Guess which distributions provide EL images for Raspberry Pis. Hint, it is Rocky Linux and Alma Linux. I bet Red Hat doesn't have a number on how many people are using RHEL because they tried it out on their Raspberry Pi first with Rocky/Alma Linux and then switched to RHEL. I'm one of them. (With the free RHEL license, for what its worth.)
So what does this mean for RHEL? I'm not an Oracle (haha!) so I can't predict the future for RHEL.
I am far from a "trend setter" and I don't know how many people have been positively influenced by my Podman coverage. Trying out Podman didn't come out of the blue. I did that by dipping my toes on Fedora first, then experimented a lot with a "production-grade" environment by using Rocky Linux and then finally deployed some of my own services on RHEL. Not to say "If I didn't do it, no one would have", but you can't deny the impact by making accessible content like this, from me and many others. FYI, I prompted Abhishek to cover Podman, not the other way around.
Sure, some of the above mentioned "contributions" may not help Red Hat, but it helps whom Red Hat serves as their customers.
So, Should We Be Worried? Yes, and No.
The decision that Red Hat has taken, does make sense as a business. But it also doesn't. It makes sense for short term goals. Not in the long term.
Red Hat, as a company that proudly contributes back to its upstream – that will never stop. Even by "soft-paywalling" RHEL code, Red Hat will continue contributing to the upstream. Red Hat will continue innovating on new developments.
They just won't keep their RHEL secret sauce “open” (in the sense that you expect it to be) anymore. And this secret sauce is nothing “proprietary” in itself. Almost everything that Red Hat offers in RHEL is open source.
Their secret sauce is with backporting patches to RHEL's stable packages. And it is a completely fair business decision to provide these patches--which are publicly available for a different version of the same package – to only RHEL customers. The task of applying patches to keep an “older version” of a package up-to-date is very difficult.
So, I do understand why they made this decision.
Red Hat is NOT close sourcing RHEL (at least technically).
Red Hat is still an excellent company that has a proven open-source portfolio. I use some of these products on a day-to-day basis like: RHEL on the Raspberry Pi 4 (yes, it's possible!), [Rootless] Podman, Cockpit, Ansible, systemd, and many more!
Here's Why You Should Worry
If RHEL isn't freely available--not free of cost, but accessible to download like Debian, Ubuntu and even Fedora; without an account – the newcomers looking at being introduced to the enterprise Linux ecosystem will keep on decreasing in numbers. I think so...
If that number goes down, the people who do actually recommend RHEL in business to use and pay for it will go down too. And you know what loop this triggers.
- Fewer newcomers interested in RHEL → Less RHEL subscriptions purchased
- Less revenue for Red Hat → Fewer Red Hat contributions to Upstreams (systemd, Podman, Linux Kernel, Guh-NOME, Wayland, Nvidia collaboration, etc.)
- Finally, → Fewer improvements to the Linux ecosystem in general
Sure, Red Hat isn't the only company contributing back to the Linux ecosystem, but you can't deny its monstrous impact on pushing the ecosystem full stream ahead!
My Response to Red Hat Statements
I'm not attacking Mike McGarth. This is just a direct response to his statements. I believe that I have the right to communicate my point of view as an RHEL user who ended up loving it so much as to deploy his personal blog via RHEL from his Raspberry Pi 4B.
Unless specified otherwise, all the following quotes are taken from this blog post.
I feel that much of the anger from our recent decision around the downstream sources comes from either those who do not want to pay for the time, effort and resources going into RHEL or those who want to repackage it for their own profit.
Yes, that's completely fair, but again, I'll harp on this point until you realize that RHEL is nothing without the community at large that has been trained on CentOS and its successors alike.
The current IT professionals include people who once were in the group which you now tag under the umbrella of “either those who do not want to pay for the time....".😒
By displaying such a stance, you are making sure that there are significantly less number of people who are introduced to the enterprise Linux, and later work with RHEL.
We have to pay the people to do that work — those passionate contributors grinding through those long hours and nights who believe in open source values. Simply repackaging the code that these individuals produce and reselling it as is, with no value added, makes the production of this open source software unsustainable. That includes critical backporting work and future features and technologies under development upstream. If that work becomes unsustainable, it will stop, and that's not good for anyone.
This is not only a perfectly reasonable take, but it is also the harsh reality. Open source is underfunded. It is also severely hard to raise funding for open source when anybody can just use your product without giving practically anything in return.
But, if the community did not get RHEL for free (before restricting the source code), it may not have been a huge success as it is today. The community explored RHEL and made it a bigger player because it is free.
I'm not asking Red Hat to be a charity and just give it away. I want Red Hat to have more than enough funding to spend it on improving upstream. But there should be a middle ground somewhere.
Again, the free subscription is not an equivalent for a RHEL clone. Sure, I have written about how to get Red Hat Enterprise Linux for free. However, it's not the same spirit. I'll quote Jeff Geerling here:
And no, please don't post "but you can use your Red Hat Developer Subscription!" I don't have to with Debian. Or Ubuntu. Or Arch. Or... you get the point.
More recently, we have determined that there isn’t value in having a downstream rebuilder.
There most definitely is! I won't repeat myself here about how the RHEL rebuilds are the entry point for nerds like me into RHEL itself.
Windows is popular only because MSFT allowed pirated versions to exist (and cheaper license keys).
Had they clamped down, no household would have spent a dime on Vista's successors. I don't mean to compare Vista--the disaster of an OS--to RHEL, but this is an analogy that most of you will relate to, grudgingly, as you recall not getting that Windows license refund.
The generally accepted position that these free rebuilds are just funnels churning out RHEL experts and turning into sales just isn’t reality. I wish we lived in that world, but it’s not how it actually plays out. Instead, we’ve found a group of users, many of whom belong to large or very large IT organizations, that want the stability, lifecycle and hardware ecosystem of RHEL without having to actually support the maintainers, engineers, writers, and many more roles that create it.
... are you sure? 😮
Sure, the conversion rate from free users to RHEL customers may not be that high, but I bet you it's not less than 30%. If you keep up with this, the amount of new customers will be less than 10% of current new “sign-ups”.
The same argument can be made for taking advantage of the free developer license for RHEL. Creating new Red Hat accounts with temporary email IDs to take advantage of the free developer subscription is a non-trivial task for those who are uninterested in paying for RHEL.
While we are talking about conversion rate, how many of these professionals just use a RHEL clone for their internal product development (to match against RHEL) but do actually provide their customers with support when they use RHEL?
How do you measure the value provided to your customers by a third party's support? You can't, it isn't something tangible that you can measure.
I'll go on a side tangent where this actual thing happened. My current [small town, local] ISP uses RHEL to authenticate [me and other] users because the company from whom he leased his Internet line, it has an authentication product ready to be deployed on
CentOS and RHEL.
Guess what they used internally to develop against RHEL? 😊
Simply rebuilding code, without adding value or changing it in any way, represents a real threat to open source companies everywhere. This is a real threat to open source, and one that has the potential to revert open source back into a hobbyist- and hackers-only activity.
So the work that RHEL clones like Rocky Linux and Alma Linux did as the following projects is not significant enough to be accounted as separate from "simply rebuilding code, without adding value or changing it in a way"?
I don't know how to rephrase this more politely so please don't take it personally. This comes from the love for RHEL, not misdirected anger at Red Hat [management].
- Rocky Linux and Alma Linux, both have images for Raspberry Pis, making it even more accessible and cheaper to test the waters before bathing in the metaphorical blissful ocean of Enterprise Linux.
- Alma Linux has a tool called Elevate built with the help of Red Hat's Leapp framework, which allows all users (this obviously includes RHEL and even Alma's "competitor", Rocky Linux) to upgrade from major versions (i.e. from 7.x to 8.x and likewise). I'm sure Red Hat's customers would love such a tool.
- Rocky Linux has a build tool called Peridot. It allows anyone to have a custom build of RHEL. This build may be based off Rocky Linux, or it can even be a company's internal RHEL clone deployed to prevent supply chain attacks.
This reply is also relevant to the following sub-quote from a prior quote:
Simply repackaging the code that these individuals produce and reselling it as is, with no value added, makes the production of this open source software unsustainable.
They are adding value, to the EL ecosystem, just not directly to RHEL.
About CentOS Stream...
CentOS Stream is a weird product. Not bad in any sense (at least in my opinion) but weird. It differs from RHEL in the following key ways, one of which is definitely a deal breaker for someone:
- It is the testing group for RHEL, which makes things easier for RHEL and SIGs (Special Interest Group) but for your use-case, it may or may not help.
- CentOS Stream doesn't use NVER naming scheme. A minor change like this can break scripts meant to work on RHEL by testing on CentOS Stream. The same can be the case for security updates as described by a user on Twitter.
- As I mentioned, security fixes come late to CentOS Stream; while there are some reasons to do that, it does not help small businesses that remain vulnerable by using RHEL clones.
Now that CentOS Stream exists, the development of RHEL is more open than ever. But, coincidentally, RHEL's own source is locked behind a soft paywall that is subject to Red Hat's EULA.
This may help Red Hat for now. But if this stance is maintained, without even taking another action against RHEL clones, the user, developer, and support community around Enterprise Linux will wither out, I believe...
I think this will end up hurting Red Hat in the long term as the community around dies out and only exhausted IT professionals (meaning: people who are forced to work with it, like with Oracle products) are left using RHEL or worse, Oracle Linux.
I'm not ranting here to call out Red Hat employees. But instead, to let you know that if this is the way you proceed, the community, which loves RHEL, will die out.
I don't want that 😔
I wish Red Hat all the luck in pushing the Linux ecosystem forward! Genuinely ❤️