HomeAboutMailing ListList Chatter /0/0 3.239.33.139

CentOS shifting to CentOS Stream

2020-12-12 by: David White
From: David White 
------------------------------------------------------
Several of you have probably already read the announcement.
Earlier this week, CentOS announced that they are shifting the focus of the
project entirely over to CentOS Stream. In effect, that means that CentOS
as we know it is going away.

See https://blog.centos.org/2020/12/future-is-centos-stream/ for the full
announcement.

Frankly, I'm surprised that no one has brought it up on this distro list
yet. :)
I might ramble a little bit here. There's been a pretty lively discussion
going on in a different mailing list I'm a part of, so I'll post here an
edited version of what I posted over there while trying to still make some
sense.

I'm interested in you all's take.

Here is a post that I read which is in *favor* of Red Hat switching to
Stream.
https://jperrin.org/blog/thoughts-on-stream/

And here is a post I read which is *not* in favor of the switch:
https://www.cyberciti.biz/linux-news/centos-linux-8-will-end-in-2021-and-shifts-focus-to-centos-stream/

I follow Cyberctl on Twitter, and there was / is a pretty interesting
discussion going on (which I was slightly a part of), on the topic:
https://twitter.com/nixcraft/status/1336348208184741888

My perspective:
I really think Red Hat / IBM is shooting themselves in the foot. I would
likely move my stuff (that is not already off of CentOS). On a side
tangent, this is slightly frustrating for me, because I've spent a lot of
time recently studying for the RHCSA exam, and I hope to take it in
January. I feel like all of the Red Hat exams will be devalued, if there is
indeed the mass exodus that I'm sort of expecting to see at this point.

I've personally been very frustrated with how slow it takes Red Hat to
patch some CVEs. I've also experienced (at least one) customer who failed
PCI compliance audits, and I confirmed that there were CVEs that were years
old that Red Hat just didn't patch, and didn't care about. As soon as I
switched that customer to the latest version of Ubuntu with some crafty
`ufw` firewall rules for ssh, the PCI audit passed.

This looks interesting: https://rockylinux.org/
It was spun off within days of the announcement, and one of the original
CentOS co-founders is helping to head up this new Rocky Linux project.

-- 
David White

=============================================================== From: Bret McHone ------------------------------------------------------ My employer is a pretty heavy CentOS shop so they are looking at it very closely. I don't think from the current discussions going on that most are in favor of Stream and would probably look to Rocky Linux. It's reminiscent of Oracle's move with Java. -B

=============================================================== From: Billy ------------------------------------------------------ Disclaimer: I work for Red Hat, yet these opinions are my own. There=E2=80=99s a set of personae that I like to understand, aka who=E2=80=99= s my audience. Each of these personas has some need for something RHEL-like. This is my und= erstanding of their needs. 1) Individual developers that need to craft something that will run on RHEL,= but they don=E2=80=99t have the time nor desire to obtain a developer subsc= ription (aka free rhel). Aka: =E2=80=9Cit=E2=80=99s too complex or too expen= sive.=E2=80=9D 2) The community project that needs a stable RHEL-like platform, but can=E2=80= =99t afford RHEL.=20 3) the non-profit, research, academic, or other =E2=80=9Ccommunal=E2=80=9D e= ntity that needs stable infrastructure to work on =E2=80=9Ccommon good.=E2=80= =9D 4) corporate IT / Line of Business: needs FAST, EASY ACCESS to infrastructur= e for lower level environments (aka CHEAP) with LITTLE RISK as part of their= dev/QA lifecycle to =E2=80=9Ceventual production.=E2=80=9D 5) Integrated Software/Hardware Vendors (ISV/IHV) creating apps, appliances,= and systems that are shipped to customers that need to contain an OS, and t= he pricing needs to be near zero per unit, additional usage for (a) developm= ent, (b) QA, and (c) certification. 6) Small / Medium sized businesses (SMB) that need a stable platform to run t= heir production workloads, yet cost efficient: OS can=E2=80=99t cost more th= an the hardware. 7) large corporate entities that decided to run production workloads via com= munity, unsupported software, while simultaneously running enterprise softwa= re =E2=80=9Cwhere it matters.=E2=80=9D There are options. So, the good news, there are programs in the work for (1) through (5). Immed= iate relief, in the form of expansion of the developer program, expanded use= cases, community project focused, and to aid in easy access for developers,= both the corporate and individual. For (6), this is being actively discussed, there=E2=80=99s a gap, there shou= ld exist a place for SMB that has a lower bar of entry. I=E2=80=99m not part= of those discussions mind you, I bring up the pain, and the response is =E2= =80=9Cyes, we=E2=80=99re working on some ideas.=E2=80=9D For (7), I=E2=80=99ll be blunt. An enterprise company, spending $100 million= a year of hardware and software shouldn=E2=80=99t be trying to save pennies= by running community software. Penny saved, pound foolish. CentOS was NEVER= RHEL. It lacked the certifications, both hardware and security (FIPS, STIG,= CC), it lacked a feed back loop (engineers support RHEL, not centos), it wa= s not built by RH engineering - not one released bit, and crucially, there w= as never a =E2=80=9Cpath to supportability=E2=80=9D for CentOS. There were 3= rd parties that would =E2=80=9Cadopt=E2=80=9D your CentOS, but you would nev= er be able to pay RH to adopt your CentOS infrastructure. There is a tool, convert2rhel, which will convert CentOS to a fully supporte= d RHEL system. I wrote an internal blog about it, but here=E2=80=99s an exte= rnal facing one [1]. I=E2=80=99d like to address the =E2=80=9CPCI certification discussion. I=E2=80= =99ve guided about 15 customers over the past 7 years through various certif= ication processes. At its core, it=E2=80=99s a relationship between the comp= any and the auditors: it=E2=80=99s never been pivoted on a single technology= issue. Typically, when the auditor brings up an issue, they ask for evidenc= e of why something is the way it is, and justification for why it can=E2=80=99= t be changed, plus any mitigation that could be applied. I bring in RHEL sec= urity engineering for these discussions. We layout our process, provide repo= rts on what we are mitigating, and justify the risk. You will never get that with CentOS, Rocky Linux, or Cloud Linux. That=E2=80= =99s only a RHEL conversation. And finally, this is a blog [2] from my good friend and colleague, Scott McC= arty aka Father Linux. It dips into some of the monetary aspects of what ope= n source means, as well as bringing a personal touch to the conversation. And another final note: Jim Perrin from the blog below, is a great person. I= have the honor of calling him a friend, and have lots of over-beer discussi= ons about CentOS over the years. He=E2=80=99s at Microsoft, now, but he=E2=80= =99s still on the CentOS Governance Board.=20 1: https://www.redhat.com/en/blog/convert2rhel-how-update-rhel-systems-place= -subscribe-rhel 2: http://crunchtools.com/before-you-get-mad-about-the-centos-stream-change-= think-about/ --b e project entirely over to CentOS Stream. In effect, that means that CentOS a= s we know it is going away. nnouncement. et. :) oing on in a different mailing list I'm a part of, so I'll post here an edit= ed version of what I posted over there while trying to still make some sense= .=20 m. hifts-focus-to-centos-stream/ ussion going on (which I was slightly a part of), on the topic: ikely move my stuff (that is not already off of CentOS). On a side tangent, t= his is slightly frustrating for me, because I've spent a lot of time recentl= y studying for the RHCSA exam, and I hope to take it in January. I feel like= all of the Red Hat exams will be devalued, if there is indeed the mass exod= us that I'm sort of expecting to see at this point.=20 ch some CVEs. I've also experienced (at least one) customer who failed PCI c= ompliance audits, and I confirmed that there were CVEs that were years old t= hat Red Hat just didn't patch, and didn't care about. As soon as I switched t= hat customer to the latest version of Ubuntu with some crafty `ufw` firewall= rules for ssh, the PCI audit passed.=20 entOS co-founders is helping to head up this new Rocky Linux project.

=============================================================== From: "Alex Smith (K4RNT)" ------------------------------------------------------ I'm gonna take a closer look at Springdale Linux. It's a RHEL-based distro released by Princeton University. " 'With the first link, the chain is forged. The first speech censured, the first thought forbidden, the first freedom denied, chains us all irrevocably.' Those words were uttered by Judge Aaron Satie as wisdom and warning... The first time any man's freedom is trodden on, we=E2=80=99re al= l damaged." - Jean-Luc Picard, quoting Judge Aaron Satie, Star Trek: TNG episode "The Drumhead" - Alex Smith - Kent, Washington (metropolitan Seattle area) =80=99s my audience. er oo expensive.=E2=80=9D =E2=80=99t =9D entity that as n or more than hould of those es, we=E2=80=99re ion a by G, ere were 3rd er be able to pay =99s an =E2=80=99ve guided about s. tors: it=E2=80=99s y r =80=99s at on. . but he=E2=80=99s -subscribe-rhel think-about/ e shifts-focus-to-centos-stream/ is rs

=============================================================== From: Stephen Kraus ------------------------------------------------------ Might as well mark the end for Scientific Linux, but IIRC they were ending active development anyways. On Sat, Dec 12, 2020 at 2:40 PM Alex Smith (K4RNT) wrote: o all =80=99s my per too expensive.=E2=80=9D =9D entity that as in for t more than should t of those yes, we=E2=80=99re lion a s by IG, here were 3rd ver be able to pay =99s an =E2=80=99ve guided about es. itors: it=E2=80=99s e ny or =E2=80=99s hat ion. soft, now, e-subscribe-rhel -think-about/ n me -shifts-focus-to-centos-stream/ f is d ars

=============================================================== From: John D ------------------------------------------------------ Thorough response, but I kind of disagree with the point on 7. There's plenty of reasons to use a community supported distro if you're an enterprise, but it depends on what kind of enterprise you are. A lot of big tech companies use centOS because they don't mind supporting everything themselves. They're going to use some deployment software, anyway. They're going to use custom repos. Hell, they may even customize the base OS. Why pay for RHEL license if you're already going to do that level of support for your OS and you have the staff to do it? (note: these companies should be returning something to the open source community imo. Either dev work or donating to open source projects and foundations that they use.) Those companies are honestly going to decide if they want to go to stream or if they're just going to go to an entirely new distro. Maybe they'll make their own community project. Who knows? It's a bit annoying but I'm pretty sure most of those companies aren't worried. I think what pisses people off the most is EOLing CentOS 8 so soon. I think that's a really bad look, tbh. Otherwise, I don't really care about this move. RHEL/CentOS can do what they want. Good, thorough response though. Appreciate your thoughts and opinions. =80=99s my audience. er oo expensive.=E2=80=9D =E2=80=99t =9D entity that as n or more than hould of those es, we=E2=80=99re ion a by G, ere were 3rd er be able to pay =99s an =E2=80=99ve guided about s. tors: it=E2=80=99s y r =80=99s at on. . but he=E2=80=99s -subscribe-rhel think-about/ e shifts-focus-to-centos-stream/ is rs

=============================================================== From: Billy ------------------------------------------------------ I=E2=80=99m not sure how Scientific Linux was built: is it converted CentOS,= or did they repackage the released source? https://centos.org/distro-faq/#q3-will-the-source-code-for-red-hat-enterpris= e-linux-continue-to-appear-on-gitcentosorg To fulfill the GPL, Red Hat releases all of our source code to git.centos.or= g. In fact, we=E2=80=99re going above and beyond, as the GPL only requires u= s to release to our customers that we=E2=80=99ve shipped our product. But we release to everyone. This allows things like centos to exists. I thought scientific Linux was also one of the distros that repackaged from t= he source. Oracle Linux being another. I=E2=80=99m sure there are several others. --b active development anyways. o released by Princeton University. he first thought forbidden, the first freedom denied, chains us all irrevoca= bly.' Those words were uttered by Judge Aaron Satie as wisdom and warning...= The first time any man's freedom is trodden on, we=E2=80=99re all damaged."= - Jean-Luc Picard, quoting Judge Aaron Satie, Star Trek: TNG episode "The D= rumhead" =99s my audience. understanding of their needs. HEL, but they don=E2=80=99t have the time nor desire to obtain a developer s= ubscription (aka free rhel). Aka: =E2=80=9Cit=E2=80=99s too complex or too e= xpensive.=E2=80=9D =E2=80=99t afford RHEL.=20 entity that needs stable infrastructure to work on =E2=80=9Ccommon good.=E2= =80=9D cture for lower level environments (aka CHEAP) with LITTLE RISK as part of t= heir dev/QA lifecycle to =E2=80=9Ceventual production.=E2=80=9D ces, and systems that are shipped to customers that need to contain an OS, a= nd the pricing needs to be near zero per unit, additional usage for (a) deve= lopment, (b) QA, and (c) certification. un their production workloads, yet cost efficient: OS can=E2=80=99t cost mor= e than the hardware. community, unsupported software, while simultaneously running enterprise so= ftware =E2=80=9Cwhere it matters.=E2=80=9D mmediate relief, in the form of expansion of the developer program, expanded= use cases, community project focused, and to aid in easy access for develop= ers, both the corporate and individual. hould exist a place for SMB that has a lower bar of entry. I=E2=80=99m not p= art of those discussions mind you, I bring up the pain, and the response is =E2= =80=9Cyes, we=E2=80=99re working on some ideas.=E2=80=9D lion a year of hardware and software shouldn=E2=80=99t be trying to save pen= nies by running community software. Penny saved, pound foolish. CentOS was N= EVER RHEL. It lacked the certifications, both hardware and security (FIPS, S= TIG, CC), it lacked a feed back loop (engineers support RHEL, not centos), i= t was not built by RH engineering - not one released bit, and crucially, the= re was never a =E2=80=9Cpath to supportability=E2=80=9D for CentOS. There we= re 3rd parties that would =E2=80=9Cadopt=E2=80=9D your CentOS, but you would= never be able to pay RH to adopt your CentOS infrastructure. orted RHEL system. I wrote an internal blog about it, but here=E2=80=99s an e= xternal facing one [1]. =E2=80=99ve guided about 15 customers over the past 7 years through various c= ertification processes. At its core, it=E2=80=99s a relationship between the= company and the auditors: it=E2=80=99s never been pivoted on a single techn= ology issue. Typically, when the auditor brings up an issue, they ask for ev= idence of why something is the way it is, and justification for why it can=E2= =80=99t be changed, plus any mitigation that could be applied. I bring in RH= EL security engineering for these discussions. We layout our process, provid= e reports on what we are mitigating, and justify the risk. =80=99s only a RHEL conversation. McCarty aka Father Linux. It dips into some of the monetary aspects of what= open source means, as well as bringing a personal touch to the conversation= . n. I have the honor of calling him a friend, and have lots of over-beer disc= ussions about CentOS over the years. He=E2=80=99s at Microsoft, now, but he=E2= =80=99s still on the CentOS Governance Board.=20 lace-subscribe-rhel nge-think-about/ : the project entirely over to CentOS Stream. In effect, that means that Cent= OS as we know it is going away. ll announcement. t yet. :) on going on in a different mailing list I'm a part of, so I'll post here an e= dited version of what I posted over there while trying to still make some se= nse.=20 ream. d-shifts-focus-to-centos-stream/ iscussion going on (which I was slightly a part of), on the topic: d likely move my stuff (that is not already off of CentOS). On a side tangen= t, this is slightly frustrating for me, because I've spent a lot of time rec= ently studying for the RHCSA exam, and I hope to take it in January. I feel l= ike all of the Red Hat exams will be devalued, if there is indeed the mass e= xodus that I'm sort of expecting to see at this point.=20 atch some CVEs. I've also experienced (at least one) customer who failed PCI= compliance audits, and I confirmed that there were CVEs that were years old= that Red Hat just didn't patch, and didn't care about. As soon as I switche= d that customer to the latest version of Ubuntu with some crafty `ufw` firew= all rules for ssh, the PCI audit passed.=20 l CentOS co-founders is helping to head up this new Rocky Linux project.

=============================================================== From: "Alex Smith (K4RNT)" ------------------------------------------------------ This is a source-compiled distro, supposed to be drop-in compatible with RHEL. It's different from Scientific Linux, which folded a while back. I have yet to try it. " 'With the first link, the chain is forged. The first speech censured, the first thought forbidden, the first freedom denied, chains us all irrevocably.' Those words were uttered by Judge Aaron Satie as wisdom and warning... The first time any man's freedom is trodden on, we=E2=80=99re al= l damaged." - Jean-Luc Picard, quoting Judge Aaron Satie, Star Trek: TNG episode "The Drumhead" - Alex Smith - Kent, Washington (metropolitan Seattle area) OS, or rise-linux-continue-to-appear-on-gitcentosorg only oduct. : g d all =80=99s my y oper too expensive.=E2=80=9D =9D entity that K as =9D ain for st more than a e s should rt of those =9Cyes, we=E2=80=99re llion a es by R TIG, t There were 3rd ever be able to pay =80=99s an I=E2=80=99ve guided about ses. ditors: it=E2=80=99s he any for e =E2=80=99s t what tion. osoft, now, ce-subscribe-rhel e-think-about/ t post l d-shifts-focus-to-centos-stream/ d of e is ed ears l

=============================================================== From: Billy ------------------------------------------------------ John, that=E2=80=99s spot on. I was going to write a follow up to express that, but I just hadn=E2=80=99t d= one it yet. So folks like Facebook, Intel, US National Labs, are using CentOS Stream. In= fact, since its release a year ago, they have been its biggest proponents A= ND contributors. And I=E2=80=99m glad for that! I=E2=80=99m in sales at Red Hat. I ultimately get paid when my customers buy= more of my stuff. But guess what? I don=E2=80=99t care if you use RHEL, or use any Red Hat products. I don=E2=80=99t care. That=E2=80=99s not my goal. My goal is to provide value, and to help your organization solve things in o= rder for you to do better at your business. It just so happens that I also believe that what I have to offer is normally= the best way to solve those problems. Now, some things I don=E2=80=99t do. S= omethings, I=E2=80=99ll ask a partner for help. I might introduce your organ= ization to App Dynamics, or I might suggest you use that Net App Trident sys= tem you have for something that might on the surface compete what I=E2=80=99= m offering. But it=E2=80=99s not a competition if it does whatever thing you= need it to do better than what I can do plus you might also already have it= . This also means I=E2=80=99m not that bothered by folks using community softw= are. As long as you (and your stakeholders) know the risks, and are willing t= o take those risks on yourself - great! If you=E2=80=99re contributing your o= wn time, money, or effort back into the community (any oss community) then e= ven better! Practitioners of open source software are the gears that make OSS run. Before I ever worked at Red Hat, I was active in the gentoo community. I hav= e patches with my name on it, I have pages of responses in the mailing lists= . I saw stuff that needed love, and I was willing to help. I=E2=80=99ve continued to participate in the community. I don=E2=80=99t run g= entoo anymore, but I=E2=80=99ve created patches for foreman, for fog-aws, fo= r pulp, and other projects. I=E2=80=99m an advocate for changes in direction= , to bump up those BZ=E2=80=99s in the open community that I know my custome= rs need. I bleed open source, so I do want to give back. And that=E2=80=99s what=E2=80=99s missing about the talk of this announcemen= t. I really wish folks would read Scott McCarty=E2=80=99s blog. He writes it= so much better than I could. CentOS stream is about giving a path for folks= to give back. CentOS Stream is not beta. It=E2=80=99s not rawhide for RHEL. It=E2=80=99s a= stable OS, its what folks like Facebook have bet the farm on, because it gi= ves them the ability to drive direction and contribute, yet still run a stab= le OS. It=E2=80=99s what Intel is using to deliver the next gen drivers, it=E2= =80=99s the hardware enablement that we=E2=80=99ve never had before as we we= ren=E2=80=99t able to induct partners like Intel into our internal processes= . Previously, it was literally a black box to those outside RH engineering. CentOS Stream changes all of that, and as one of my colleagues said: =E2=80=9C= it=E2=80=99s made RHEL radically transparent.=E2=80=9D If consumers want to hop to another distro, that=E2=80=99s great! If they se= e no value in RHEL, and want to take on the risks themselves, then they SHOU= LD seek out a distro that meets their needs. =46rom my sales perspective, folks that were using CentOS, because of its RH= EL security and stability, yet didn=E2=80=99t want to PAY for RHEL, won=E2=80= =99t magically become customers and start paying for RHEL today. RH essentia= lly lost a $0 sale. That=E2=80=99s ok. I also know customers won=E2=80=99t b= e calling Rocky Linux to request they patch a new CVE, or add some hardware e= nablement for some card that came out last week, and they sure won=E2=80=99t= be asking ME to adopt and in-place support their 4,000 installs of Rocky Li= nux. And I=E2=80=99m fine with that. Folks pay for stuff they perceive has value,= and if you don=E2=80=99t see the value in it, then use something else. What I do care passionately about is if OUR OWN PROCESS AND MODEL pushed my c= ustomer to use CentOS for dev/Qa/test purposes, and now they need a solution= . I hear that loud and clear. I want to help, and I=E2=80=99m fighting to ge= t that out ASAP. This also applies to ISV/IHV, and embedded folks, or folks t= rying to containerized their app for distribution. I want to help, and I=E2=80= =99m not going to leave my customers high and dry.=20 --b nty of reasons to use a community supported distro if you're an enterprise, b= ut it depends on what kind of enterprise you are. verything themselves. They're going to use some deployment software, anyway.= They're going to use custom repos. Hell, they may even customize the base O= S. Why pay for RHEL license if you're already going to do that level of supp= ort for your OS and you have the staff to do it? (note: these companies shou= ld be returning something to the open source community imo. Either dev work o= r donating to open source projects and foundations that they use.) r if they're just going to go to an entirely new distro. Maybe they'll make t= heir own community project. Who knows? It's a bit annoying but I'm pretty su= re most of those companies aren't worried. I think what pisses people off th= e most is EOLing CentOS 8 so soon. I think that's a really bad look, tbh. Ot= herwise, I don't really care about this move. RHEL/CentOS can do what they w= ant. =99s my audience. nderstanding of their needs. EL, but they don=E2=80=99t have the time nor desire to obtain a developer su= bscription (aka free rhel). Aka: =E2=80=9Cit=E2=80=99s too complex or too ex= pensive.=E2=80=9D =80=99t afford RHEL.=20 entity that needs stable infrastructure to work on =E2=80=9Ccommon good.=E2= =80=9D ture for lower level environments (aka CHEAP) with LITTLE RISK as part of th= eir dev/QA lifecycle to =E2=80=9Ceventual production.=E2=80=9D es, and systems that are shipped to customers that need to contain an OS, an= d the pricing needs to be near zero per unit, additional usage for (a) devel= opment, (b) QA, and (c) certification. un their production workloads, yet cost efficient: OS can=E2=80=99t cost mor= e than the hardware. ommunity, unsupported software, while simultaneously running enterprise soft= ware =E2=80=9Cwhere it matters.=E2=80=9D mediate relief, in the form of expansion of the developer program, expanded u= se cases, community project focused, and to aid in easy access for developer= s, both the corporate and individual. hould exist a place for SMB that has a lower bar of entry. I=E2=80=99m not p= art of those discussions mind you, I bring up the pain, and the response is =E2= =80=9Cyes, we=E2=80=99re working on some ideas.=E2=80=9D ion a year of hardware and software shouldn=E2=80=99t be trying to save penn= ies by running community software. Penny saved, pound foolish. CentOS was NE= VER RHEL. It lacked the certifications, both hardware and security (FIPS, ST= IG, CC), it lacked a feed back loop (engineers support RHEL, not centos), it= was not built by RH engineering - not one released bit, and crucially, ther= e was never a =E2=80=9Cpath to supportability=E2=80=9D for CentOS. There wer= e 3rd parties that would =E2=80=9Cadopt=E2=80=9D your CentOS, but you would n= ever be able to pay RH to adopt your CentOS infrastructure. rted RHEL system. I wrote an internal blog about it, but here=E2=80=99s an e= xternal facing one [1]. =80=99ve guided about 15 customers over the past 7 years through various cer= tification processes. At its core, it=E2=80=99s a relationship between the c= ompany and the auditors: it=E2=80=99s never been pivoted on a single technol= ogy issue. Typically, when the auditor brings up an issue, they ask for evid= ence of why something is the way it is, and justification for why it can=E2=80= =99t be changed, plus any mitigation that could be applied. I bring in RHEL s= ecurity engineering for these discussions. We layout our process, provide re= ports on what we are mitigating, and justify the risk. =80=99s only a RHEL conversation. cCarty aka Father Linux. It dips into some of the monetary aspects of what o= pen source means, as well as bringing a personal touch to the conversation. . I have the honor of calling him a friend, and have lots of over-beer discu= ssions about CentOS over the years. He=E2=80=99s at Microsoft, now, but he=E2= =80=99s still on the CentOS Governance Board.=20 ace-subscribe-rhel ge-think-about/ he project entirely over to CentOS Stream. In effect, that means that CentOS= as we know it is going away. l announcement. yet. :) n going on in a different mailing list I'm a part of, so I'll post here an e= dited version of what I posted over there while trying to still make some se= nse.=20 eam. -shifts-focus-to-centos-stream/ scussion going on (which I was slightly a part of), on the topic: likely move my stuff (that is not already off of CentOS). On a side tangent= , this is slightly frustrating for me, because I've spent a lot of time rece= ntly studying for the RHCSA exam, and I hope to take it in January. I feel l= ike all of the Red Hat exams will be devalued, if there is indeed the mass e= xodus that I'm sort of expecting to see at this point.=20 atch some CVEs. I've also experienced (at least one) customer who failed PCI= compliance audits, and I confirmed that there were CVEs that were years old= that Red Hat just didn't patch, and didn't care about. As soon as I switche= d that customer to the latest version of Ubuntu with some crafty `ufw` firew= all rules for ssh, the PCI audit passed.=20 CentOS co-founders is helping to head up this new Rocky Linux project.