20 thoughts on - How Will Fragmentation Help Red Hat

• Competition in the Enterprise Linux space is a good thing. If a company or community other than Red Hat starts serving a market that RHEL can’t, it forces Red Hat to evaluate and adjust. It keeps everyone pushing and developing solutions that hopefully benefit end users and customers. If everyone is fully participating in open source and upstream, it makes them all better inherently.

  Is it different from the non-Enterprise Linux ecosystem? What do you do there given the large variety of Linux distributions?

  My personal take on this is to think about what I use and why I use it. How does something solve my needs? Does it need to be better?
  etc.

  For example, long before I ever worked at Red Hat I was a Fedora Linux user. I love that project and distribution. I literally owe my career in some part to it. In recent years, I don’t use Fedora heavily. Partly because of my day job, but also partly because my personal needs changed. I do still install almost every release in some way and try it out though. If someone asked me for a recommendation on a community Linux distribution, it would still be at the top of my list. Not because of what it was like in the past, but because of what Fedora is today which is far better than it ever has been.

  If someone asked me for a recommendation on an Enterprise Linux operating system, I’d say RHEL. Yes of course because I work on it, but also because I firmly believe it is the best on the market. It’s what I run on my main machine every day. If someone asked for a community Enterprise Linux project, I’d suggest CentOS Stream because of the direct ties to RHEL, but also because I believe it’s a relatively young and growing project with a lot of potential to do really interesting things. However, I would probably ask what their needs were and then I’d earnestly try to make a recommendation based on that.

  It is. It’s difficult for an individual to decide, and it’s difficult for a project or company to continuously push themselves to make sure they are the best option for the broadest number of users.

  Respectfully, I don’t think it’s new. We’ve had RHEL, SLES, OEL, CentOS Linux and Ubuntu for more than a decade. Rocky, Alma, whatever SUSE’s new RHEL fork is, etc are certainly newer but the situation itself is not new. I see it as an expansion of options, but the same set of considerations still applies. Which distribution and community aligns best with your needs, goals, and beliefs? Which one would you tell your friend to use?

  For me, it’s still Fedora, CentOS Stream, and RHEL.

  josh

• IMHO, there are insider politically correct opinions about the recent changes and then, there are the individual opinions of community members, end-users and the general public.

  IMHO, if you work for RedHat (IBM) your opinion could be slightly biased because of your career.

  But, the history of open source is full of examples of what happens when corporations try to create commodities from distributions backed by support contracts.

  IBM wants to make money, PERIOD. They paid billions for RedHat and investors, executives, want ROI and profit, period. No excuses.

  So, they are locking down RedHat and closing channels to important software/materials. It is what companies do all the time.

  I predict a decline in sales, a decrease in subscriptions and a percentage of the community moving away from Fedora / CentOS.

  It’s only logical reaction.

  Does IBM deserve to make a profit for buying RedHat? Yes, indeed.

  However, this is not the best way, it is the same mentality of Microsoft, Oracle and others whose products are EASILY replaced and out performed by open source community software.

  IBM has had many successes over the years, many first innovations, but also a history of mistakes and flops too! This is a flop.

  —–Original Message—–
  From: Josh Boyer
  Reply-To: CentOS mailing list
  To: CentOS mailing list
  Subject: Re: [CentOS] How will fragmentation help Red Hat Date: Thu, 13 Jul 2023 07:20:50 -0400

  Competition in the Enterprise Linux space is a good thing.  If a company or community other than Red Hat starts serving a market that RHEL can’t, it forces Red Hat to evaluate and adjust.  It keeps everyone pushing and developing solutions that hopefully benefit end users and customers.  If everyone is fully participating in open source and upstream, it makes them all better inherently.

  Is it different from the non-Enterprise Linux ecosystem?  What do you do there given the large variety of Linux distributions?

  My personal take on this is to think about what I use and why I use it.  How does something solve my needs?  Does it need to be better?
  etc.

  For example, long before I ever worked at Red Hat I was a Fedora Linux user.  I love that project and distribution.  I literally owe my career in some part to it.  In recent years, I don’t use Fedora heavily.  Partly because of my day job, but also partly because my personal needs changed.  I do still install almost every release in some way and try it out though.  If someone asked me for a recommendation on a community Linux distribution, it would still be at the top of my list.  Not because of what it was like in the past, but because of what Fedora is today which is far better than it ever has been.

  If someone asked me for a recommendation on an Enterprise Linux operating system, I’d say RHEL.  Yes of course because I work on it, but also because I firmly believe it is the best on the market.  It’s what I run on my main machine every day.  If someone asked for a community Enterprise Linux project, I’d suggest CentOS Stream because of the direct ties to RHEL, but also because I believe it’s a relatively young and growing project with a lot of potential to do really interesting things.  However, I would probably ask what their needs were and then I’d earnestly try to make a recommendation based on that.

  It is.  It’s difficult for an individual to decide, and it’s difficult for a project or company to continuously push themselves to make sure they are the best option for the broadest number of users.

  Respectfully, I don’t think it’s new.  We’ve had RHEL, SLES, OEL, CentOS Linux and Ubuntu for more than a decade.  Rocky, Alma, whatever SUSE’s new RHEL fork is, etc are certainly newer but the situation itself is not new.  I see it as an expansion of options, but the same set of considerations still applies.  Which distribution and community aligns best with your needs, goals, and beliefs?  Which one would you tell your friend to use?

  For me, it’s still Fedora, CentOS Stream, and RHEL.

  josh

• I agree and acknowledge there is bias for employees. Most of us try very hard to be aware of it and think critically through it, but it does exist.

  I will politely point out that your implication that IBM had direct input into Red Hat’s recent announcements is an assumption on your part and not based on facts.

  I’m failing to see how this email helps further the conversation that Simon started in earnest. I think Simon asks good questions and it’s worth a discussion. If your suggestion is to not recommend Red Hat distributions, what would you recommend instead and why?

  josh

• I think I finally need to remove myself from the CentOS mail list but coming from @redhat worker trying to explain what their company has done, is pretty disingenuous to say the least. It’s pretty clear what they are trying to do and it’s all driven by greed, have seen it over and over in the opensource world, and lets be clear thats what it always comes down to is greed. Bottom line its a d*ck move by Redhat, but technically still meets the letter of the GPL so it is what it is. IMHO they basically became another Oracle and we know how most feel about them, but hey someone has this great idea to make more money they have the right based on the GPL to do it..I moved a few servers to Rocky when they killed CentOS but this is it for me, I will migrate my remaining servers over to anything but Redhat, they are dead to me. I had been using CentOS for many years, when Karanbir Singh was running things and they would go to meet ups and you could get t-shirts etc..Was a great run but Redhat has ruined all that and now I
  just could care less what Redhat does from here on out. I’m nobody, but where I do work we have options for which linux distro that we want to run, I can assure you I will not be spinning up an Redhat instances…fool me once, fool me twice…

  Tom Bishop

• I have to agree with Tom Bishop. IBM has completely destroyed Red Hat and it’s clones for me. Debian GNU Linux is the next natural step. It will probably be a mix of Debian GNU Linux, FreeBSD and OpenBSD used. Let’s face the facts, corporate entities have never been the friends of the Open Source Community and they do nothing but take and control. It’s time for a change.

  -Salim

• Tom nailed it for me, similar boat. We moved to Rocky as a holding position with their LTS, but it just doesn’t make sense any more to use a Redhat base with any new server.

  The trust I think has gone since the CentOS rug pull, and that’s important. I do think we’ll see some changes to come (I can’t work out if Suse forking RH is interesting or not), but I think most people want stability with security.

  I would like to say a genuine thanks to all those who have helped previously.

  Ian

  CentOS mailing list CentOS@CentOS.org https://lists.CentOS.org/mailman/listinfo/CentOS

• Am 13.07.23 um 14:42 schrieb Tom Bishop:

  Well, as RH’s announcement is quite some day ago, I had time to reflect this jumble. The whole thing is much more complex than people want to admit and I will not decompose this all here now. Honestly I see the open source ecosystem like a hardware store. You have everything that you need to build your own home, thats all. So, some entity is needed to build it – a worker, consultant, hobby crafts(wo)man, agency, midsize firm, corporation et cetera, and that is the truth the we all should face it. To make it clear, what product do you get when a loosy community build a distribution, with components of projects that are under financed? You need to put energy into something to keep it alive, this does not a happen magically. The how can be discussed. IMO, it should not be about the content (code, its already open) its should be about the structure … something that balances the input with the output to stay sustainable. It should also be recognized that RH has contributed and continues to contribute much. I say this without any affiliation to RH – just have a large window of time available to oversee it all.

• I’ve been trying to figure out what SUSE meant when they announced a
  “hard fork” of RHEL.  If they mean to maintain a fork that remains interface-compatible with RHEL (and a fork that doesn’t remain compatible doesn’t make much sense, because the thing that everyone wants is the benefit of RHEL’s integration with other vendors), then they’ll probably periodically branch from Stream, the same way that RHEL
  does.

  If that happens, and if it’s successful, the irony is that through poor communication, Red Hat might have actually made progress in creating a
  *less* fragmented ecosystem of distributions all conforming to a common ABI, descending from Stream.

  I think that’s exactly the opposite of the direction that Red Hat is moving.  CentOS Stream makes the RHEL product more open than it has ever been — including making it easier to create real Enterprise-ready products that compete on level ground with RHEL, in ways that clones never could.

  https://medium.com/@gordon.messmer/in-favor-of-CentOS-stream-e5a8a43bdcf8

• I see this hypothesis relatively often, but there’s no evidence to support it, and it doesn’t make much logical sense.

  If Red Hat (or IBM) wanted to “lock down” RHEL, they wouldn’t be focusing on Stream, which opens it up (and makes it easier to fork.)

• I get it, I have seen you post this over and over in the mail list in regards to how great CentOS Stream is. Well its not what the original CentOS was, its something very different, it may work for you and others but it’s not the same. If it were we would not be having this discussion and there would be no need for Rocky, alma etc..If it was what the community wanted then we would not have seen the mass exodus to other distro’s that came to try to fill the gap, no it’s not the same at all, it’s different and thats fine but folks are free to chose to use it or something else. Seems most have chosen to use something else and RH continues to burn them down also…again it’s there choice, so now that I see what they really want to do I also have a choice to not to continue or recommend their product, again as referenced by one of the many RH articles, we are all just freeloaders so we shouldn’t be missed.

• IMHO Red Hat is focusing on Stream, because the community helps them to build RHEL, without losing their own benefit of the long term support in the final product, RHEL.

  Regards, Simon

• You’ll get distributions like Debian, Arch or also FreeBSD and other BSDs. One thing they had in common with Red Hat distributions is that they are of high quality and one could fully trust them.

  Unfortunately I fail to still trust Red Hat as I did in the past.

  Regards, Simon

• I don’t believe there are any Red Hat articles that call user freeloaders.

  I’m aware of one personal blog, not on the redhat.com site, in which the term “freeloader” is discussed as a term sometimes used informally by staff to describe large businesses that run production networks on a self-supported distribution, but issues that affect that production network are reported through a small number of support subscriptions. 
  That’s a very narrow use, and it’s specifically organizations who abuse the support that Red Hat sells, by proxying production issues through RHEL systems.

  And it’s the behavior of individuals, not the position of the company.

• Kudos to everyone’s input and lots of intelligent talking points!

  After decades of working in enterprise, Fortune 100 or less, I always struggled as the disruptor, as the open source evangelist, trying to get decision makers to embrace open source solutions.

  Years later, I see that commercial linux like RedHat & SuSE make a critical mistake in packaging linux as a commodity with licensing and subscriptions.

  If you think about it, the most common argument against open source is always support & expertise. THAT IS WHERE THE MONEY IS………

  Selling licenses and subscriptions is par for the course for IBM
  because they have doing that model for decades. But, many enterprise decision makers are STILL afraid of open source precisely due to perceived expertise and support.

  If RedHat gave away the OS and shared everything with the community and then followed up with consulting engagements and the current support engineering already in place, I think the customer base would start growing immediately.

  Open Source distributions are not designed to be licensed software in a box like Windows. Never the intention, ask Stallman if you know him.

  The money is in support and consulting. Teach the customer how to fish and they feed themselves and the footprint keeps growing!

  —–Original Message—

• Eh I beg to differ it’s pretty clear that Magnus is calling folks freeloaders… https://www.linkedin.com/pulse/problem-rocky-linux-free-beer-magnus-glantz/

  But I know he is just a principal solution architect so I know it’s just his opinion, but no it’s crystal clear where RH stands in regards to this ..

• I’ve read that article several times, including just now with an eye out for such an accusation, and I just don’t see it.

  (But, again, that is a blog from an individual employee, and not the position of the company.)

• Thank you for this response, Josh. I can sympathize with your love of Fedora, and it is nice to see how you characterize CentOS Stream. I, too, hope Stream evolves and grows into something that’s not just a place for contributions, but also something that draws users and encourages innovation.

  I am NOT a Red Hat employee (few things are more obvious than that). From the outside, I don’t know if inside Red Hat there is a culture at war with itself, or if everybody is of one mind. That wouldn’t mirror any free software community I’ve seen, so I’d like to think there’s some (hopefully)
  healthy debate going on.

  Right now I feel like CentOS Stream is not where it’s “meant to be,” and nobody seems to be talking about exactly what that is. I know what I want it to be, but I’m not sure the project itself has figured out its direction.

  I’m thinking the downstreams that base on Stream will be doing the things that users want and Stream itself isn’t doing right now. That’s everything from the live media that AlmaLinux provides (a great thing!) to the critical patches that Stream doesn’t send to users in a timely manner.

  I do want to thank all those Red Hat employees and other community members who believe in free software and community distributions.

• —————————————–From: “Steven Rosenberg”

  To: “CentOS mailing list”
  Cc:
  Sent: Sunday July 23 2023 5:13:08PM
  Subject: Re: [CentOS] How will fragmentation help Red Hat

• Somehow the text of my message got deleted:

  I’m still using CentOS 7 because both security and stability are important to me. While CentOS Stream may, or may not, have comparable security, it is severely lacking in stability. IMHO both CentOS 8 and CentOS 9 are gigantic piles of garbage.

  When the time comes that CentOS 7 no longer meets my needs, I’ll be switching to some other distro, most likely Debian.

  JP

  —————————————–From: jefflpost@twc.com To: “CentOS mailing list”
  Cc:
  Sent: Sunday July 23 2023 6:29:52PM
  Subject: Re: [CentOS] How will fragmentation help Red Hat

  —————————————–From: “Steven Rosenberg”

  To: “CentOS mailing list”
  Cc:
  Sent: Sunday July 23 2023 5:13:08PM
  Subject: Re: [CentOS] How will fragmentation help Red Hat

• +1

  I now have only two servers left on CentOS – both on 7, as 8 and 9 and the whole stream thing just never passed my confidence tests. All my new machines since stream was enabled have been on Ubuntu LTS. It has been a learning curve, a little more work to deal with certain updates and the
  5 years vs 10 years of CentOS is going to be more work still – however they at least provide an OS upgrade path that seems to work just fine, although apps and config changes are required as versions update.

  Will keep watching, but having worked in large corporate environments with IBM as a major IT partner, there is no way I’m holding my breath for any meaningful change.

  My thanks to the wonderful team that made CentOS what it was (up to version 7), and I understand their reasons to embrace the change – just not something I’m prepared to live with.

  Shalom Rob