Hacker News Reader

Comment by freedomben | Hacker News Reader

Disclaimer: I work for Red Hat but I'm here entirely with my own opinion

My God, it's been a long time (if ever) since I've seen so much misinformation out there. I mostly blame Red Hat for the very poor delivery, but tech people are not immune to the tendency for dramatizing, misrepresenting, and polarizing. Please stop spreading this. CentOS is NOT dead. CentOS is changing, but not all that much (yes, really).

I'll do my best to answer questions here, but I'm also going to rush a blog post to try and get it out ASAP that corrects a lot of this.

Edit: Here's the blog post. Excuse typos and other things, I rushed this ;-)

https://freedomben.medium.com/centos-is-not-dead-please-stop...

View on HN

I appreciate the blog. I was pretty alarmed by the announcement, and I think as you mention a lot of it boiled down to that devilish-sounding line:

> If you are using CentOS Linux 8 in a production environment, and are concerned that CentOS Stream will not meet your needs, we encourage you to contact Red Hat about options

But, after reading your message I can partially see your point of view, in particular your gist seems to be that in the chain of development, CentOS Stream will sit where RHEL used to be, thus if you believed in RHEL as a boring, stable, reliable OS in the old world then so too should CentOS Stream be a boring, stable, reliable OS in the new one.

This logic makes sense, with one exception: this assumes that the standard for release quality remains unhindered pre/post swapping the position of RHEL / CentOS.

Before, Red Hat was incentivized to make sure RHEL releases were solid gold because paying customers were on the receiving end. CentOS releases were downstream of solid gold, making CentOS solid gold as well.

Now that the positions are reversed, what incentive does Red Hat have to make sure things are stable and low-risk before shipping to CentOS Stream? This is ultimately a matter of trust: if CentOS Stream gets a bad release, no paid enterprise server explodes. Quality no longer aligns with profit motive, a terrifying insight.

Thanks for the response, and thanks for reading. You make a terrific point:

> This logic makes sense, with one exception: this assumes that the standard for release quality remains unhindered pre/post swapping the position of RHEL / CentOS.

I don't really have an answer for that. It does seem like it would come down to "trust" and as a skeptical person I don't like that very much.

As I think about it more, while we have realigned some economic incentives for the better, it does seem (in addition to your points) there might also be an "incentive" to break CentOS from time to time, just to remind enterprises who are using it that "RHEL" is the gold standard and CentOS is "unsupported" D-: Knowing Red Hat and many of the volunteers that work on these projects, I would be shocked (and indeed would quit my job immediately) if such a thing were to happen, although it would be hard to prove.

That said such a thing would no doubt backfire. There would be outrage of course, and it would hurt the CentOS brand (which also hurts RHEL since these days the two are inextricably linked). We all fully expect another distro to "take the place" of old CentOS as a bug-for-bug rebuild (I'm excited about Rocky Linux personally) so not will there still be an option for that, but competition is great for quality of projects :-D

Anyway, much appreciate the comment :-)

I'll keep mulling it over as well.

>>it would hurt the CentOS brand

That ship has sailed. Even if what you say is true, the public reaction this is change has tarnished all the Brands, CentOS, RHEL, RedHat, and IBM...

It will be a very long time to rebuild the trust, and honestly if RedHat stays the course and refuses to support CentOS 8 for the full life cycle I think CentOS will usage (and RHEL usage) will drop off a cliff

I have already seen several software vendors with RHEL only official support announcing they are working on debian and ubuntu certification, and since it seems CentOS Stream at the most will only be supports for a 5 year life cycle down from 10, I am not sure why I would not move to debian at this point. debian is just as stable, the one thing CentOS had over it was the 10 year support, that is gone now

This still seems a little sensational. Fedora would be the explosion here. If Stream is a release candidate they still have every reason to keep it stable so RHEL updates more quickly.

> This logic makes sense, with one exception: this assumes that the standard for release quality remains unhindered pre/post swapping the position of RHEL / CentOS.

There's another issue, which I haven't seen talked about so far. With CentOS downstream from RHEL, there were the Release Notes: before updating from CentOS 8.x to CentOS 8.y, you could read the CentOS release notes (and the voluminous RHEL release notes corresponding to it) to see if there was any breakage to expect for your use case. For the security updates within a release, you could follow https://lists.centos.org/pipermail/centos-announce/ (sadly non-working for CentOS 8) and/or the voluminous RHEL errata pages corresponding to it.

With Fedora, before every update I glance at https://admin.fedoraproject.org/updates/ to see if there is any known issue with the update I'm about to do; this has saved me more than once (for instance, there was a broken selinux-policy update which would relabel the whole /home not long ago). I recall Debian had something similar (which shows you any release-critical bug recently filed for the packages you're about to install). Will CentOS Stream have something like that? If I run "dnf update" and see it's about to update bash, is there anywhere I can look to see what changed, and whether someone complained about the update?

Great point also. I am going to pass your question to some people that I hope can answer it.

I got a response. It looks like this is not currently implemented, but some details are still being worked out so it's possible we could get something like that. A digest has been requested before and thought is going into what it would look like and what any technical hurdles would be.

I also learned a new command that I'm adding to my toolbox:

    dnf changelog --upgrades

That probably depends, how strongly Stream and RHEL actually are going to be linked. If this relation is indeed similar to current situation, but centos and RHEL changing places, then there isn't much point holding QA until RHEL release. Actually, there is going to be good initiative: when you upstream is unusable, because too buggy and untested, then what? Whole point of Stream seems to be usable base to RHEL.

> but not all that much (yes, really).

Cutting 8 years(I think? have something like 2029 in the back of my head) of support from the thing a bunch of people just moved to, with no publicized plan what they'll have to move to instead in the next months(?), doesn't sound like "not all that much" for the kind of environments I've seen CentOS in. What's the plan for them?

This is actually the big area where I think the criticisms are right and fair. Dropping 8 years was uncool.

Moving to Stream is an option immediately (see post[1] for more information about how CentOS is NOT unstable or "the new Fedora" or whatever else everybody is saying), but there are also forthcoming announcements about new free RHEL subs for individuals and edu institutions and the like. I don't know anything about those. I personally think it was terrible to announce the CentOS changes without those to accompany, but hey if I ran things everything would be perfect :-D

[1]: https://freedomben.medium.com/centos-is-not-dead-please-stop...

Don't care. Nothing I run will ever run a red hat product again, and most of the people I've talked to feel similar.

Thanks. Luckily not my personal problem (or if it is nobody has told me yet...), but knowing how much hassle goes into any major upgrade (even if it's smooth sailing technically) in places with overly conservative processes, I hope those paths and changes get clearly communicated.

Totally agree. This has the potential to be a nightmare of epic proportions. I strongly wish Red Hat hadn't done that.

At the same time if you take any piece of free (as in beer and speech) software and don't pay a vendor promising support, you are kind of taking a risk that whoever is doing work for you for free might decide to stop. In no way does that take Red Hat off the hook here, but it is something I consider.

I've got a non trivial migration myself to handle before next December. Depending on how well Stream goes in my testing (so far it's been flawless but I'm not totally done with testing yet) and depending on what RH announces next month (or two), I may be cursing their names next fall ;-)

For me, with a single production server, migrating to Ubuntu LTS on a new server seems a much less risky proposition to me than an in-place upgrade to either RHEL or Stream.

Forgive me, but if you're just building out a new server, wouldn't it be easier and less risky to stick with RHEL or Stream since you already know your workload runs fine? You also already know all the packages you need and probably have config files all set up and everything. At a minimum I would think they would pretty equal as far risk and ease if you're rebuilding.

My workloads are mostly containerized at this point so the host OS is pretty insignificant overall. Perhaps you're in the same boat, although I'd still think it's at worst equal, at best easier to stick with RHEL flavored.

Canonical has not, to my knowledge, ever dropped 8 years of support off a release - even the non-paid ones! Redhat has. In some people's minds, that changed the risk assessment significantly; not "well it worked on this RHEL-family OS, so it should work on this other RHEL-family OS", but "I need to replace my OS - should I use one from the same vendor that just screwed me over, or should I switch to literally anything else?" Honestly, you've got people seriously suggesting Oracle Linux over anything from Redhat; do you have any idea how much trust you have to burn to get people to treat Oracle as a good option!?

The majority of my service is a compiled C++ server with html/js front-end, so it runs on practically anything. I do need to run ffmpeg/firefox/pulseaudio to record sessions, which has been problematic on Ubuntu in the past, but it seems to work reliably on Ubuntu 20. I also have many customers running my service on their Ubuntu servers, so it shouldn't be an issue.

From what I've seen, Ubuntu 20 is probably going to be simpler to set up than Centos 8, as (I think) it has many of the required packages such as ffmpeg without needing to add epel and various other third party repos. It also doesn't have selinux, which I generally need to turn off on Centos (after spending a lot of time trying unsuccessfully to get everything working properly with it).

My only issue with Ubuntu LTS is the shorter 5-year lifecycle.

Perhaps by this time next year there will be some alternative repo that we can simply switch to in Centos 8 so that it just continues as-is, without us needing to take the risk of upgrading everything to a completely new distro.

> there are also forthcoming announcements about new free RHEL subs for individuals and edu institutions and the like.

This is the surprising part. It seems like customer service 101 to have those announcements ready when this announcement was made. By not having those details ready in a timely fashion, they inherently devalue those programs.

> the thing a bunch of people just moved to

Yep, if this had been for 7, it would’ve been just “meh” and move on to Arch BTW or something. But I just reinstalled new servers to move to 8 and now that was a waste of time.

> Cutting 8 years(I think? have something like 2029 in the back of my head)

Correct: https://web.archive.org/web/20201101131417/https://wiki.cent...

It’s a bit arrogant telling everyone that they’re wrong when you’re changing the entire purpose of a product. It’s dead.

It was accurately described already in the press release and that killed it.

If there is backtracking it will make it more dead as it would break trust.

It’s dead. You killed it some more here by shouting down the community.

Edit: to clarify I have over 100 CentOS boxes which will not be running CentOS or RHEL next year.

> Edit: to clarify I have over 100 CentOS boxes which will not be running CentOS or RHEL next year.

What distro(s) are you thinking of migrating towards?

Ubuntu LTS at the moment.

Why is that better than Debian in your opinion? We also run Ubuntu LTS but I need ammo.

It's not better. It's just less scary for the CxO class.

I'm sure other commenters will put it much better than I could, but for those people that use CentOS for the purpose of a stable long term release, surely it is dead. The distribution now being called "CentOS" is almost at the opposite ends of the scale in terms of stability. The fact they've called this other thing the same name is purely marketing.

If you destroy something then create something entirely different with the same name then the fact remains that you still destroyed the original thing. That's not dramatising or hyperbole.

> The distribution now being called "CentOS" is almost at the opposite ends of the scale in terms of stability.

No, this is totally wrong (which is why I'm trying to correct it). CentOS will be where RHEL used to be. It is trading places with RHEL. It's not "the opposite end of the scale in terms of stability." CentOS will be roughly a single minor point release ahead of RHEL.

Please read more about because that is super misleading and others are seeing things like this and propagating them: https://freedomben.medium.com/centos-is-not-dead-please-stop...

Saying this repeatedly does nothing to make it actually true.

I agree! Saying it's false repeatedly likewise doesn't make it false.

Where am I wrong? Can you point out my error and make a case for your position? What makes you think RHEL is still "stable" but suddenly a distro a tiny bit ahead is now wildly unstable?

Go be a child somewhere else. Your responses show a clear lack of understanding of the point of using CentOS and becoming petulant when someone calls you on it just makes it worse.

CentOS was binary compatible because the point was expecting it to be the same environment for application layer and testing. The change being described throws all that in the trash and its intellectually dishonest at best or just obtuse ignorance at worst to "misunderstand" the difference here, and how big of a change switching to a "Beta" effect release schedule is when there was an existing 10 year commitment to release schedule that was also thrown in the trash.

Don't try to insult the userbase by then gaslighting after the fact.

Don't you understand. He made a picture and swapped places between RHEL and CentOS. Clearly this proves that CentOS Stream is replacing RHEL.

It’s swapping places with RHEL beta, not RHEL. Not calling it a beta doesn’t not make it a beta. It’s directly taking the role that RHEL beta releases used to have (ie. be a point release ahead, and be for the public to test and report issues before going to stable RHEL).

Regardless of what your blog post says, RHEL was not a beta for old CentOS. Old CentOS was a rebuild/copy of RHEL, bugs and all included. RHEL will not be a copy/rebuild CentOS Stream, so it’s not a reversal of roles.

Using outdated packages doesn’t make a distribution stable. It’s stable when its combination of packages and configuration has already been tried by a bunch of people, and confirmed to work, or has had its issues thoroughly documented with effective workarounds.

Rather than "stable long term release," I would phrase it as a "dot release frozen in time as much as possible modulo serious bugs and security fixes." Which is not the general direction that software is headed in. Changes going into CentOS Stream will have gone through QE tests. So, yes, it is different but, increasingly, most of the software you use will be using this type of model. Certainly SaaSs don't let you choose a dot release for the most part.

> Rather than "stable long term release," I would phrase it as a "dot release frozen in time as much as possible given serious bugs and security fixes."

Which is exactly what we want.

> So, yes, it is different but, increasingly, most of the software you use will be using this type of model. Certainly SaaSs don't let you choose a dot release for the most part.

Not all of us are happy with having to follow the update treadmill. Software is a tool, like an electric drill; one doesn't update their electric drill every day, otherwise we would be wasting all our time updating and dealing with the fallout of updates, instead of drilling holes in wood. If you develop (or do anything) on top of software, you want a stable base, instead of one shifting all the time; one builds castles in rock, not in quicksand.

Absolutely fair that you desire that, but it's not where the software world in general is headed and it's at least understandable that software companies are skating to where the puck is headed. (ADDED: Of course, with open source, you have no requirement to upgrade. It's just that folks want to make certain upgrades--like security fixes--and for serious bugs but no others.)

The nice thing about open source--rather than cloud SaaSs--is that there are more backward-looking options from a variety of sources. I'd also encourage users who need or feel they need that sort of option to contribute to such efforts so that they're sustainable.

> Which is not the general direction that software is headed in.

You could be right. But "we got rid of this thing (and gave something totally different the same name) because that's not the direction the software is heading" would be a different comment from the one I replied to, which was "this thing still exists, honest!"

Sure. Since you say this is not a big deal and CentOS is not dead but is just changing, I really only have 2 questions:

1) What in your opinion were the main use cases for CentOS before this change announcement?

2) How do you think CentOS Stream is still viable for those use cases?

Great questions, thank you. It may take me a bit to answer, but I will :-)

Ok ready to answer.

1) I obviously can't speak for anyone but myself here, but in my use case, which I think is pretty common, is to use CentOS for production deployments of various web apps and services where a person prefers the rock solid stability of RHEL without incurring the cost, and does not need support to go along with it.

2) I explain this in depth in this blog post which is up now ( https://freedomben.medium.com/centos-is-not-dead-please-stop... ), but the short version is that CentOS Stream is basically going where RHEL used to be. If RHEL was too bleeding edge for you, then CentOS Stream will be too, but otherwise it will be approximately the same as RHEL used to be. I consider RHEL and CentOS to be pretty equal, with CentOS lagging a bit behind. Now they will be pretty equal, with RHEL lagging a bit behind. Packages going to CentOS are packages that would have gone directly to RHEL previously.

> I consider RHEL and CentOS to be pretty equal, with CentOS lagging a bit behind. Now they will be pretty equal, with RHEL lagging a bit behind. Packages going to CentOS are packages that would have gone directly to RHEL previously.

There will be another small difference: the number of updates. Just for fun, I started a centos:8 image in podman, installed the package which switches it to centos streams, and ran an update (which by the way partially failed with an error message, IIRC it tried to change /proc in a way podman doesn't like). Then I looked at the version of one of the updated packages before and after the update (IIRC it was bash), and looked at its changelog. There were several releases between these two versions, and I would expect each one of them to appear separately as an update within CentOS Stream.

That is, packages that would have gone directly to RHEL previously will be going to CentOS, but not all packages going to CentOS will be going to RHEL; we will also be getting all the intermediate steps.

It's clear and literally self-described as a "development branch". To have anyone try to spin that as more stable and suitable for production is disingenuous.

> Before stuff goes to CentOS Stream it has passed RHEL QA and CI.

Do we have an official word on that? Because it feels like QA could just as easily be priority tasked on the next point release RHEL and starve stream of QA while it carries on ahead. I guess this boils down to the workflow which there doesnt seem to be a simple guide of with this news.

And no, I wouldn't agree that they just swapped places. CentOS stream is a rolling release. Unless I am mistaken that's not how RHEL is built right now.

I can see a reason for swapping them in that adding RHEL goodies/licencing into the free version is easier than removing, but the release cycle change, dropping CentOS 8 and as you admit, absolutely shambolic official press release, with lines that reek of a salesperson trying to squeeze out a bonus, do not inspire faith.

I see that you note in your profile that you work at Red Hat. May I ask in which department of Red Hat? The reason I ask that is, are you part of the decision making team which made this decision? If not, then no matter how good you think the intentions of Red Hat management here are, I can assure you that they haven't been made for the betterment of the CentOS community. If this was anything to do with the betterment of the community then this wouldn't have been a sudden announcement but instead would have been a long drawn discussion involving the community. I don't say this just for the sake of it, but I have closely followed how they killed the JBoss application server community and the project, as I note in my other comment[1] and this whole CentOS announcement is exactly on similar lines. I have known people in the application server team, who initially tried to justify the whole process, just like you are doing it now for this CentOS "migration", but even they have now realized there wasn't any truth to it.

It's not a co-incidence or an accident that the messaging around the community side of this CentOS announcement is confusing and unclear (they are letting the community members try and decipher what this all means for the CentOS project) instead it is intentional.

The bottomline, just like in that other case, is that the sales team at Red Hat cannot justify the hundreds of thousands of dollars they ask for supporting RHEL, when customers (or prospective customers) note that they can get the same quality from CentOS (stable release, regular bug fixes, community help).

[1] https://news.ycombinator.com/item?id=25358847

Sorry but yes CentOS is dead

CentOS, is COMMUNITY ENTERPRISE OPERATING SYSTEM

The community is dead, the Enterprise part of the Operating system is dead

Sure the CentOS name will be still around, but is clear that CentOS is now a Dev focused operating system not designed or targeted for Stable Enterprise operations with a community of Enterprise SysAdmins around it

If you believe the changes are "not all the much" then you are clearly not in touch with the user base of CentOS, outside of Devs that is

The changes are GREAT for devs, and Appliance Manufacturers. Maybe even for Cloud customers

But for us Old School, OnPrem sysadmins running CentOS for Stability, and Long Term Support, CentOS is most certainly dead

> is clear that CentOS is now a Dev focused operating system not designed or targeted for Stable Enterprise operations with a community of Enterprise SysAdmins around it

This is absolutely not true. Would you have described RHEL this way? If not, then you shouldn't describe CentOS this way either. CentOS took RHEL's place in the chain. Instead of Cent being downstream of RHEL, the two were swapped around.

> If you believe the changes are "not all the much" then you are clearly not in touch with the user base of CentOS, outside of Devs that is

Say what you will. I've been using Cent since 2012 and currently run numerous workloads in production on it. I've shipped SaaS products on top of it and have been responsible for widespread maintenance. I've been a package maintainer for numerous packages as well.

> But for us Old School, OnPrem sysadmins running CentOS for Stability, and Long Term Support, CentOS is most certainly dead

I know plenty of old school on prem sysadmins who run CentOS for stability and LTS (I'm one of them. I have 3 bare metal machines in prod at the moment running CentOS). I may go with Rocky for one of those machines (it's a router/load balancer), I haven't fully decided yet.

That said I agree with you somewhat, that is the one area where this change could be somewhat negative. However, with CentOS being week and months behind RHEL right now, do you see that as a good thing? Do you like having an unpatched system for days or weeks at a time when an embargoed security fix hits RHEL and has to wait for CentOS to build and distribute? I don't. Security patches are top priority for me.

Edit: Blog post mentioned is now live: https://freedomben.medium.com/centos-is-not-dead-please-stop...

Easy problem to solve.

https://linux.oracle.com/switch/centos/

Similar to forcing Microsoft or Google to prompt the user for their desired browser or search provider, it would be interesting if rhel was forced to prompt for a repo provider.

It would have been wiser to splash rhel support upgrades on CentOS gdm or cli cockpit than to threaten to kill a product.

Really, what were you thinking? Nazgul indeed! Fork you!

Old habits die hard.

Ha! Red Hat is too evil so you move to Oracle :‑D

> Really, what were you thinking? Nazgul indeed! Fork you!

Glad we were able to have a productive conversation

Oracle offered freedom from the perspective of rhel, and I used it.

I actually was granted IPO shares because of the things I wrote two decades ago.

I think there was a fundamental misjudgment in what you have done. Oracle is more expensive on most tiers, but offers a $100 support level (forum-only, with updates) that undercuts all rhel agreements.

Had you offered this support level, and advocated it (as I said) with gdm and CLI advertisements, you might have kept a great many in the fold.

You did not do this, and now Oracle updated their repo conversion for CentOS 8.

It is not too late. You should reconsider.

>>However, with CentOS being week and months behind RHEL right now, do you see that as a good thing? Do you like having an unpatched system for days or weeks at a time when an embargoed security fix hits RHEL and has to wait for CentOS to build and distribute? I don't. Security patches are top priority for me.

This is absolutely a False dilemma fallacy, there is absolutely nothing technically or legally prohibiting RedHat from releasing patches for both at the same time. Nothing. It is a sales/marketing or a internal process choice to do it that way.

Changes to do this process, even moving CentOS to be ahead of RHEL is not the problem. The "rolling release" or non-release, or what every model they are calling it now with CentOS Stream is the problem. The fact that it will no longer be Binary Complete with RHEL is the problem, the fact that CentOS will be used as a "beta" (and spare me the marketing BS claiming it is not that) is the problem

> Would you have described RHEL this way?

RHEL is a targeted, Released Enterprise Operating system with a 10 year support cycle

CentOS was a targeted, Released Enterprise Operating system with a 10 year support cycle

CentOS Stream is not a targeted, released Enterprise operating system with 10 year support cycle. CentOS Stream is pseudo-rolling release with no point releases, used to build a targeted, released enterprise operating system.

> RHEL is a targeted, Released Enterprise Operating system with a 10 year support cycle

> CentOS was a targeted, Released Enterprise Operating system with a 10 year support cycle

And you don't see the problem here? Why do you expect Red Hat to maintain a free OS which targets exactly the same use case as their paid OS? They are not a charity.

It would be great to have a free alternative to RHEL, but you shouldn't feel entitled to Red Hat providing that for you.

> Why do you expect Red Hat to maintain a free OS which targets exactly the same use case as their paid OS?

Because they said they would.

They could have made the change effective prior to releasing CentOS 8, or they could have made it effective for all versions after CentOS 8. Either of those paths would be understandable. Making the change after many have already migrated to 8 sends the message that Red Hat commitments are not to be trusted.

Well first and foremost because the software is open source released under the GPL so yes to any code that is released under the GPL I am entitled to it for free, that is the terms of the license and the fundamental promise of FOSS software

Secondly I do not believe I stated anywhere that I was entitled to it, this seems to be a strawman you created in some attempted to project your own beliefs of my motivations

In several comments I have stated my biggest issue with RetHat/CentOS announcement was not they were shifting focus, but rather that they did so while reducing their previously stated committed of supporting CentOS 8 until 2029, to 2021.

This is a betrayal of trust, and I do believe people are entitled to hold companies to account, and demand they honor their public commits of which RedHat previously committed to supporting CentOS 8 until 2029, and I feel they should honor that commitment. If they want to discontinue the support or shift the focus of the project for future releases then more power to them, the community will take the GPL code as we are entitled and create a replacement CentOS.

> the fact that CentOS will be used as a "beta" (and spare me the marketing BS claiming it is not that) is the problem

If you just want to vent, go ahead. I'm not going to engage with someone assuming bad faith. If you actually care to hear the argument about why saying "beta" is stupid, here it is: https://freedomben.medium.com/centos-is-not-dead-please-stop...

> CentOS will no longer be downstream of RHEL as it was previously. CentOS will now be upstream of the next RHEL minor release. This is NOT a “RHEL Beta”

That's... what beta means.

EDIT: Let me reframe it this way: If CentOS Stream is just becoming RHEL, then why does RHEL exist? If it's truly just as good as the old RHEL was, then explicitly merge them - either give RHEL away 100% freely like CentOS and charge for support, or discontinue the RHEL brand and sell support for CentOS. The fact that you have a downstream RHEL says that that is more stable, because that is what you're willing to sell for money.

EDIT2: Actually, thinking it over more and trying to be charitable, I can see a way for that to make sense: If you believe that CentOS is moving to RHEL's spot and RHEL is getting even more stable/slow/tested, then you can say that CentOS isn't getting any worse and ... sort of object to calling it RHEL beta. Except that it still is RHEL-beta, just moved up the scale of stablity.

I don't work for Red Hat, but I saw a post that was saying that there is something in the works to fill the niche that CentOS offered. It just wasn't ready yet so the announcement hasn't been made yet. However, they needed to send a statement out regarding CentOS so companies were aware of what would be happening with CentOS 8 and perhaps avoid migration right now.

For a community that depends on LTS, the sysadmins seem to make some rash decisions/drama. I think the logical decision here would surely be to wait and see what happens and make a decision in a year when support ending would actually become relevant.

Some of this is timing.

Many many many companies, right or wrong, wait until the very list minute to upgrade their legacy systems.

Many of us are either still in the process or worse just finish our migration from CentOS 6 to CentOS 8, only to have the rug pulled out from under us.

The sheer impact this move will have on many organizations is what is driving this emotional reaction

If they would have honored the LifeCycle for CentOS 8, there would not have been the reaction you have seen

100% of the emotion / drama is coming from the choice RedHat has made to cut the supported life of CentOS 8 by 7 years. that is a DRASTIC action, and their belief that 12 months is "enough time" shows a clear ignorance of how the product is used in enterprise or they fully understand and are banking on the desperation of enterprise to sell RHEL licenses.

either way it drastic choice by RedHat.

Well, we mostly agree! I agree with everything until:

> their belief that 12 months is "enough time" shows a clear ignorance of how the product is used in enterprise or they fully understand and are banking on the desperation of enterprise to sell RHEL licenses.

There's another explanation, and I think it makes a little more sense than yours which would require Red Hat, literally the company whose core job it is (which they are damn good at, enough so that you use it) to provide an enterprise distribution to some of the most conservative companies on the planet, to not understand how the product is used in enterprise.

The suggestion that RH is "banking on the desperation of enterprise to sell RHEL licenses" at least makes sense given the facts, and that's kind of what I thought too when I first saw the news (the way they worded it sure made it seem like this was a money grab).

It is possible however (and indeed this is my belief) that Red Hat doesn't consider Stream to be that big of a change, and they've given a full year to migrate. Indeed if you read my blog post[1] I don't think it's that big of a deal.

I have zero inside info on this but I would bet highly that sales of RHEL were a factor in this decision for sure. I don't see how they couldn't be. But given how long Red Hat has been providing all their sources publicly (which they DON'T have to do) and acquiring and open sourcing companies, you don't give them even a tiny benefit of the doubt here that maybe their not just one-sided evil capitalists trying to squeeze nickels out of the CentOS community for short term gains?

[1]: https://freedomben.medium.com/centos-is-not-dead-please-stop...

I have not read your blog post yet, I will however I do want to comment on this

If you claim RedHat understands the CentOS Client base so well, they how did they not predict and address all of the concerns that we OnPrem Enterprise Admins have? How did they fail at the messaging soo bad, and why has there been no formal statements addressing the issues we have had?

Why has the CentOS Site not been updated to clearly address that "CentOS Stream is not that big of a change you guys are all over reacting"

I suspect because it is a pretty big change, and they know it

Ok so I have read your blog post, I would like to address some of the things you raise

First you ask "Did you think of RHEL as a beta for CentOS, if not you should not think of CentOs Stream as beta for RHEL"

This is flawed in a big way, CentOS was complied from RHEL sources, only having RedHat trademarks removed. It was binary for binary compatible. RHEL 8.3 is the EXACT SAME as CentOS 8.3, so not it was not a beta.

This can not be said for CentOS Stream, first off there are no point releases, no releases at all for CentOS Stream, it is a Continuous Stream of updates (thus the name) for the entire 5 year support window (down from 10 year support for CentOS). So where one could plan and update from point release to point release there is no such way to do that with CentOS Stream, that type of release cadence is not acceptable for a production work load

Further Red Hat has explicitly said CentOS Stream is to be used "for the development of RHEL" so how else can one take that, the updates applied to CentOS Stream will then be used to roll up into a point relase of RHEL, sure seems like beta in all but name to me... Sure maybe it will be more stable than other types of Beta Testing, and maybe the baggage that comes with the word beta is a bit to strong for this case. Still CentOS Stream is not binary compatible with a Point release of RHEL that much is clear, to many of use that would classify it as a Beta

Secondly you also claim that RedHat or CentOS has acknowledged that is botched the messaging and it is water under the bridge. I have been following this issue pretty closely, and I do not see any such acknowledgement. I suppose it could have been on social media, i do not follow social media, but it certainly was not added to the Blog Post, nor has a new blog post been issued acknowledging it, nor has the original blog post been updated to clarify things

So no I dont think it has been acknowledged, nor it is water under the bridge we should move on from

Third you talk about "CentOS being old crusty etc", for many of us this is the EXACT reason we use CentOS, we support long lived legacy system that we do not want to change very often, some of use run software are a DECADES old, some dating back to long before I was born (and I am a middle aged man). We need, want a "crusty old" STABLE system. I am not runing the latest NodeJS project, or Ruby, or Rust, or Go, or what ever the hot new startup lang of the week is today...

I am not running docker, k8s, or any of the "hot new" technologies.

I am running an old EPR, an old database, an old Java app, or some other legacy system. I want stability, not hot new feature. CentOS is my choice for stability, if I want to play in the wild world of new hotness I have Arch for that.

as to licensing sales, RHEL shots them selves in the foot here. For example tagging their no support license as "not for production use" is a bad move IMO, why is not for production use, if I do not need RHEL support, why should not use a RHEL licensed system in production?

There are many many other problems with RHEL sales but that is for another day.

Though I do agree shoving CentOS in the face of a RHEL sales person is a shit thing to do. I hope it does not happen all that often, that said if a RedHat sales person does not have ready answer for that well that should be day 1 training for RedHat sales staff.

One final thing, the publication of sources. Some of the code they absolutely are required to release the sources. any Code that GPL. Now I understand not all of RHEL is GPL, and some if it would get in the grey area of what is "linking" code, but that is for lawyers to debate over, and like you I hope the day never comes where the lawyer need to debate over that issue

RedHat has done alot of good in the community, but FOSS is a three leg stool. Trust is one of those legs, and RedHat is taking a hacksaw to that leg right now....

Completely agree. I am critical of Red Hat for not announcing this all at the same time, but I likewise very much agree with you. We still have a full year and they've said to expect an announcement within a month or two (at least that's what I read elsewhere, nothing I've heard internally).

Still have a full year? You think replatforming is something that happens in a day?

“We aren’t going to fulfill our LTS support promises of CentOS 8, but just wait a few months until we announce the replacement. Trust us, it will be different this time.”

Thanks for the blog post. Ep383 of the Linux Unplugged podcast [1] also went into detail with CentOS members.

What CentOS gave you until Stream was (with some delay) the "same" set of packages that RedHat released in RHEL, which was (hopefully exhaustively) tested by RedHat QA, including making sure that packages interact in a sane way. That's what earned RHEL (and by extension CentOS) that designation of stability. By virtue of being upstream, CentOS Stream could have a lot more updates (the kind that would previously be internal to RedHat). I doubt same level of QA can/will be invested for each update state (if it could be, RHEL would probably be a rolling distro). If you don't have a QA safety net of your own, the risk is higher with Stream vs 8.

You could hack point releases into Stream: wait for RHEL release, then dnf versionlock Stream packages to the same release versions as RHEL (assuming you can map versions of modular packages).

[1] https://linuxunplugged.com/383

If you have great ability to correct misinformation, then I would like you to bring back Red Hat 9 from 2003, fully updated with the latest RPMs.

Red Hat demonstrated then exactly what it intended to do now.

> CentOS is changing, but not all that much

The lie detector said that is a lie.

The 'not all that much' is too much for most

Good luck with that position that your customers are wrong and you didn't just fuck them all over (even the paying ones). You might want to update your linkedin.

We call that "shooting the messenger." I didn't do anything to anyone lol. I was as surprised as everybody else when I saw the news (and I know you didn't read my blog post now btw because I discuss that in there).

If shitting on some person on the internet makes you feel better, by all means take it out on me.

Reply on news.ycombinator.com