"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."
"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."
===
This decision is good for Red Hat's bottom line, but bad for the RHEL ecosystem long term.
This decision restricts RHEL so that it is effectively closed source. Some businesses and governments will continue to use RHEL for the time being, but most users have other Linux distributions to choose from.
Most users and developers won't use RHEL or develop software against it because RHEL's use is restricted to paying customers. EPEL will shrink and only have software that big businesses, governments, and similar organizations need. In the long run, this means that RHEL will be less like other Linux distributions and more like AIX or Solaris.
The way commercial enterprises are set up favors them exploiting the current state of their ecosystem to the point of self-destruction. This outcome matters not to shareholders, as they can jump ship accordingly.
This is a serious flaw in our economic system, with extremely dire long-term consequences if unaddressed. Yet it seems surprisingly hard for people to wrap their head around it.
Examples abound. Google and their search, journalism, open discussion forums and other social platforms...just to name a few. And of course in physical space, environmental destruction of all sorts goes on unimpeded.
IBM’s business model is more about squeezing existing customers rather than finding new ones. They made the calculation that squeezing the existing RedHat base is more important than expanding it. This is what mature companies that can’t grow organically do.
>Question: Why is RHEL so popular in enterprise environments?
The answer is not CentOS. It's being early to the market, and incumbency. It's that RHEL was (and probably still is) by far the most serious enterprise Linux for a long time. It's popular in enterprise environments because it was popular in enterprises back in the day. Anything to do with CentOS is a functional irrelevancy to most of the big customers.
I worked for a healthcare IT company that did a lot of work in Ubuntu.
We discovered this - that there were two types of customers - the ones who did not care what OS we used underlying our software and ecosystem, and the ones who did.
For the ones who did not care, well, good. For the ones that did care, not a single one cared in a "we want Ubuntu" way, everyone said "we want Red Hat".
The support side of things was always the argument, and still is. We know that for many it's a crutch, a safety net, people can argue all day long about how useful support can be, but in the end, it's still a very large checkmark, and Red Hat gets that checkmark every time.
> It's popular in enterprise environments because it was popular in enterprises back in the day.
And so were Solaris and IRIX and HP/UX, until they weren't. And that happened in large part because Red Hat Linux was accessible to tech enthusiasts.
Not many could justify a Sun Ultra 1 or SGI Indigo for their homelab, but they could download Red Hat Linux for free and install it on an old PC. Eventually, these people began to question the price/performance of proprietary UNIX and hardware for their projects at work. The combination of lower price, widespread expertise, and commercial support from Linux vendors (especially Red Hat) circa 2000 is what marked the beginning of the end for proprietary UNIX.
And even when Red Hat Linux gave way to RHEL, it was still far less expensive than what the legacy UNIX vendors offered, plus clones such as Whitebox Enterprise Linux or CentOS were readily available for people to keep their RHEL skills current at home. That's helped to keep it going, but once IBM makes RHEL inaccessible to enthusiasts, the decline of RHEL in the marketplace will surely follow.
And they can still download CentOS Stream for free and install it in their homelabs. In other words, exactly what you described enabled Red Hat to succeed.
The fact that so many people have expressed concern about recent changes by Red Hat suggests to me that the company has failed to communicate effectively and/or many people no longer trust Red Hat.
What an odd comparison, and ignoring the context of the issue at hand.
Both a failure of communication in how RH have worded their stance, and breach of a social contract (The shortening of the CentOS 8 support mid-cycle that kicked this off) have caused the current tension.
The failure to communicate, matters not about truth or "lies" with whatever you seem to be implying with your comparison.
If falsehoods are interpretable from what was said, they have poorly communicated. Doubly so if they know their relationship with the target audience is already strained by past behaviour.
In this age of social media and 'extremely online' people, even the smallest misunderstanding can be magnified into an uproar, especially if it's in the interests of certain people to put pressure on RH to behave in a way that benefits these people. It's really convenient to confidently state, with no knowledge of their internal constraints and decisions, that RH screwed up. It's harder to hold off on passing judgment until all the facts are clear.
No, they can not. Before they killed CentOS you could get the support from community. Now they're killing what's left of the enthusiast community and people stop using it because nobody believes them any-more.
I was under the impression that CentOS mostly competed with Debian or Ubuntu, with the added benefit that switching from CentOS to RHEL was very easy.
RHEL probably competes more with proprietary Unix systems (although they all seem to be fading to oblivion these days), SLES, or Windows.
Maybe z/OS on mainframes, but I suspect Linux deployments on those machines are more like "We already have the mainframe, let's buy a couple more CPUs and RAM and put Linux in an extra LPAR or VM instead of buying new servers".
when I say IBM is probably not involved in this decision. They may be, but I don't think so.
IBM mostly staying out of Red Hat business decisions might have been true while folks like Jim Whitehurst and Paul Cormier, members of the old guard, were still in leadership positions, but I don't think it's true anymore.
Over the 4 years since the acquisition, I have heard from many Red Hatters that things have continuously eroded, day after day.
And is that very surprising? Something like this was inevitable, given IBMs track record with other acquisitions.
Same here. It's anecdotal but I have a decent number of colleagues who ended up at Red Hat over the years. For the first year after the acquisition they were excited and saying things like "IBM is going to change its culture to be more like Red Hat". A year or two after, they stopped saying that. And now a lot of them are actively looking to move on because they don't want to work in IBM's culture.
I don't think the average Red Hat customer will base their business decision on HN sentiment.
Also, they're probably using software certified for Red Hat Linux and don't have the freedom to pick another OS if they want the support they're paying for.
None of the suits cared about Linux in the early days when we sysadmins were extolling it's benefits. The idea that Linux could ever take on commercial Unix offerings from companies like IBM was laughable. It's so strange to me that the argument that suits don't care what geeks think is still being made. Current users will stick with Red Hat until they fear they won't be able to hire for it. But that's what happened to AIX and why IBM bought Red Hat in the first place.
Linux did not win on technical merits back then (it wasn't very good). It did not win because its geeky simps pushed it. It won because it was cheaper, good enough, and (more importantly) another company (Red Hat!, Caldera) was willing to walk you through compliance, offer support, and take the blame if shit went sideways.
Maybe your experience was true for your company. But it feels like revisionism to me, a fellow 90s Linux professional who introduced it to many orgs.
I was there when Linux overtook proprietary UNIX in the late 1990s / early 2000s and agree completely.
The Solaris and HP/UX stalwarts who laughed Linux off as a toy for hobbyists changed their tune pretty quickly when Oracle announced support for their flagship database on Linux. Adoption quickly accelerated as other software vendors announced support. Obviously Red Hat was the winner in all of this, but they could easily lose their lead if another Linux vendor gets serious.
This is simply not true: GPL requires you to share all code needed to build the binary you shipped including their minor patches. GPL also has a clause: "You may not impose any further restrictions on the recipients' exercise of the rights granted herein.".
The people the binary is shipped to can access it via weather means red hat provides. If red hat does not provide you a binary, they do not need to give you access to their code.
Yeah Red Hat does share all the code for customers who purchase binaries and they are not imposing further restrictions on recipients of those binaries with their EULA. If the EULA is violated Red Hat is simply exercising their rights to not distribute future binaries and source thus not violating the GPL clause you mention
And yet they promised to use blackmailing tactics to severe relationship with those customers who want to exercise their rights to the code and republish it.
Turns out you are right, things have changed in the SUSE world. And it does indeed seem to be worth considering. However, in terms of long term support (LTS), Leap still can't match RHEL clones.
I've had discussions with the powers that be at my work environment. I frankly think this will be the dog biting the hand and (as previously mentioned given both Redhat and IBMs greed and track record) is overdue.
I wouldn't be surprised that the community would some how get enough momentum and make a similar move that Microsoft did back in the 80s, when everyone released their own version of DOS, to go in a separate direction and turn the market.
it may take a year or two, but i wouldn't be surprised to see a new linux alternative to rhel on the horizon
There are no shortage of alternatives: SLES, Oracle, and Ubuntu if you want support. Debian, Arch, Slackware, and a host of others if you don’t care about support.
There’s even alternatives to Linux itself: the BSDs, Solaris/Illumos, even Windows Server.
I think Red Hat will stick around. This move doesn’t change much for the typical people and organizations that choose to pay for RHEL.
This comment strikes me like a breath of fresh air amongst a lot of bad breath.
RHEL has taken a step that looks like Solaris etc. of the past, but aren’t the ecosystems completely different nowadays? It seems like it’s comparing apples to oranges.
It's like slower moving Arch, but instead of being freely assembled it's a community mostly ruled by red hatters and a bunch of feel good corporate speak.
From my own company, I can definitively say that even if Alma/Rocky don't figure out a way to continue providing quick updates, we certainly won't start paying for RHEL. Discussion is ongoing right now between Debian, SuSE, and Ubuntu Server.
Distributions don't much _matter_ these days since everything is containerized anyway. Nobody's particularly upset, just annoyed that RH is causing us extra work for no particular discernible reason.
Now if IBM/RedHat closed Ansible, hoooo boy, that would start a 6 alarm fire.
44 comments
[ 3.2 ms ] story [ 99.3 ms ] thread===
This decision is good for Red Hat's bottom line, but bad for the RHEL ecosystem long term.
This decision restricts RHEL so that it is effectively closed source. Some businesses and governments will continue to use RHEL for the time being, but most users have other Linux distributions to choose from.
Most users and developers won't use RHEL or develop software against it because RHEL's use is restricted to paying customers. EPEL will shrink and only have software that big businesses, governments, and similar organizations need. In the long run, this means that RHEL will be less like other Linux distributions and more like AIX or Solaris.
This is a serious flaw in our economic system, with extremely dire long-term consequences if unaddressed. Yet it seems surprisingly hard for people to wrap their head around it.
Examples abound. Google and their search, journalism, open discussion forums and other social platforms...just to name a few. And of course in physical space, environmental destruction of all sorts goes on unimpeded.
The answer is not CentOS. It's being early to the market, and incumbency. It's that RHEL was (and probably still is) by far the most serious enterprise Linux for a long time. It's popular in enterprise environments because it was popular in enterprises back in the day. Anything to do with CentOS is a functional irrelevancy to most of the big customers.
We discovered this - that there were two types of customers - the ones who did not care what OS we used underlying our software and ecosystem, and the ones who did.
For the ones who did not care, well, good. For the ones that did care, not a single one cared in a "we want Ubuntu" way, everyone said "we want Red Hat".
The support side of things was always the argument, and still is. We know that for many it's a crutch, a safety net, people can argue all day long about how useful support can be, but in the end, it's still a very large checkmark, and Red Hat gets that checkmark every time.
And so were Solaris and IRIX and HP/UX, until they weren't. And that happened in large part because Red Hat Linux was accessible to tech enthusiasts.
Not many could justify a Sun Ultra 1 or SGI Indigo for their homelab, but they could download Red Hat Linux for free and install it on an old PC. Eventually, these people began to question the price/performance of proprietary UNIX and hardware for their projects at work. The combination of lower price, widespread expertise, and commercial support from Linux vendors (especially Red Hat) circa 2000 is what marked the beginning of the end for proprietary UNIX.
And even when Red Hat Linux gave way to RHEL, it was still far less expensive than what the legacy UNIX vendors offered, plus clones such as Whitebox Enterprise Linux or CentOS were readily available for people to keep their RHEL skills current at home. That's helped to keep it going, but once IBM makes RHEL inaccessible to enthusiasts, the decline of RHEL in the marketplace will surely follow.
Both a failure of communication in how RH have worded their stance, and breach of a social contract (The shortening of the CentOS 8 support mid-cycle that kicked this off) have caused the current tension.
The failure to communicate, matters not about truth or "lies" with whatever you seem to be implying with your comparison.
If falsehoods are interpretable from what was said, they have poorly communicated. Doubly so if they know their relationship with the target audience is already strained by past behaviour.
This is factually incorrect. At least read my comment properly before replying.
RHEL probably competes more with proprietary Unix systems (although they all seem to be fading to oblivion these days), SLES, or Windows.
Maybe z/OS on mainframes, but I suspect Linux deployments on those machines are more like "We already have the mainframe, let's buy a couple more CPUs and RAM and put Linux in an extra LPAR or VM instead of buying new servers".
Over the 4 years since the acquisition, I have heard from many Red Hatters that things have continuously eroded, day after day.
And is that very surprising? Something like this was inevitable, given IBMs track record with other acquisitions.
Nothing gold can stay.
Red-Hat, Ubuntu and SuSE.
This will hardly change it.
I've used all three of those distros at home, but now I don't.
It’s an abusive relationship.
Let them go.
Paywalling FOSS code is unforgivable and should end them. So end them.
Stop using RHEL.
Maybe your experience was true for your company. But it feels like revisionism to me, a fellow 90s Linux professional who introduced it to many orgs.
The Solaris and HP/UX stalwarts who laughed Linux off as a toy for hobbyists changed their tune pretty quickly when Oracle announced support for their flagship database on Linux. Adoption quickly accelerated as other software vendors announced support. Obviously Red Hat was the winner in all of this, but they could easily lose their lead if another Linux vendor gets serious.
https://www.theregister.com/2021/06/04/opensuse_leaps_to_153...
I wouldn't be surprised that the community would some how get enough momentum and make a similar move that Microsoft did back in the 80s, when everyone released their own version of DOS, to go in a separate direction and turn the market.
it may take a year or two, but i wouldn't be surprised to see a new linux alternative to rhel on the horizon
There’s even alternatives to Linux itself: the BSDs, Solaris/Illumos, even Windows Server.
I think Red Hat will stick around. This move doesn’t change much for the typical people and organizations that choose to pay for RHEL.
RHEL has taken a step that looks like Solaris etc. of the past, but aren’t the ecosystems completely different nowadays? It seems like it’s comparing apples to oranges.
For those of who have not dabbled in the RH space for a long time: Where does Fedora stand in this brave new world?
Distributions don't much _matter_ these days since everything is containerized anyway. Nobody's particularly upset, just annoyed that RH is causing us extra work for no particular discernible reason.
Now if IBM/RedHat closed Ansible, hoooo boy, that would start a 6 alarm fire.