CentOS replacement?OpenSUSE? Ubu20?
Michael L
Mark J. Bailey
Why not Rocky Linux (https://rockylinux.org/) from the original CentOS guy (akin to the situation with MySQL/MariaDB)? LOL, while I appreciate the reason for the name “Rocky”, I can’t help but think of the conjured mental comparison to something like “Lemon Autos”. Apparently, I’m not alone in this regard! :)
Anyways, a work colleague of mine down in Huntsville (who works in gov’t contracting) says some of his colleagues have been evaluating it as a CentOS replacement, and the results are looking up. I’ve loaded it, but not played with it much yet. And, of course, I can’t help but wonder if it will it eventually get “CentOS’d(ead)” by IBM as well. But the claim is that is (mostly?) 100% plug-n-play with RHEL8/CentOS8, but I can’t yet vouch for that personally. And to be honest, it’s been since early this year that I last dug into it.
At the end of the day, save for something like Debian core, we all are subject to some entity’s commercial backers’ objectives that may or may not line of with our long-term hopes, CentOS being a perfect example, and possibly Ubuntu as well at some point.
My $0.02.
Sent from Mail for Windows
--
You received this message because you are subscribed to the Google Groups "NLUG" group.
To post to this group, send email to nlug...@googlegroups.com
To unsubscribe from this group, send email to nlug-talk+...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/nlug-talk?hl=en
---
You received this message because you are subscribed to the Google Groups "NLUG" group.
To unsubscribe from this group and stop receiving emails from it, send an email to nlug-talk+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/nlug-talk/CALdmzXbbGuUest%2B18p2AqiRuV5azK-8f%3D%3D9xOb_pM8bdcS7OWQ%40mail.gmail.com.
--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.
Kent Perrier
And, of course, I can’t help but wonder if it will it eventually get “CentOS’d(ead)” by IBM as well.
Tilghman Lesher
killed is through the lack of resources, be it computing power,
storage, volunteer time, etc. A company comes in and provides those
resources -- for some sort of consideration, and over time, it takes
over, because they have the resources the project needs.
That's what happened to CentOS, and it will likely happen to Rocky
Linux, too. Not immediately, of course, but over time. It may get 5
or 10 years, before being subsumed -- and perhaps the next company
will also be bought out by IBM and merged into Red Hat.
If you want the project to keep going, your best bet is to contribute
on an ongoing basis, either with money or your time. Packages need to
be maintained (usually by scripting the replacement of Red Hat
trademarks with community terms), and packages need to be rebuilt
using computing resources. Commit some community effort into keeping
it going, and you'll in turn ensure that the community project won't
need outside resources by a business which might eventually take over.
Money is also needed to pay for things like bandwidth, storage, and
potentially dedicated machine(s) for the project.
> --
> You received this message because you are subscribed to the Google Groups "NLUG" group.
> To post to this group, send email to nlug...@googlegroups.com
> To unsubscribe from this group, send email to nlug-talk+...@googlegroups.com
> For more options, visit this group at http://groups.google.com/group/nlug-talk?hl=en
>
> ---
> You received this message because you are subscribed to the Google Groups "NLUG" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to nlug-talk+...@googlegroups.com.
--
Tilghman
Mark J. Bailey
In theory (litigated, or not), yes. But, should IBM cut off access to the SRPMS, whose got the money to take them to court over it? Very few other commercial entities out there depend on Red Hat’s SRPMS, so there’d likely be little incentive to lay down the consider costs to go after them. Without those SRPMS, maintaining a free Rocky or Alma (or even having a CentOS in the first place) would be next to impossible, at least in the sense of mirroring RHEL. Funny how this aspect is often glazed over. Red Hat could also share source in a way that makes it much more painful to even bother with it.
--
You received this message because you are subscribed to the Google Groups "NLUG" group.
To post to this group, send email to nlug...@googlegroups.com
To unsubscribe from this group, send email to nlug-talk+...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/nlug-talk?hl=en
---
You received this message because you are subscribed to the Google Groups "NLUG" group.
To unsubscribe from this group and stop receiving emails from it, send an email to nlug-talk+...@googlegroups.com.
Michael L
To view this discussion on the web visit https://groups.google.com/d/msgid/nlug-talk/CAHPkZcWb%3D%3DT%2BO27mQz_Mwq%2BNqshyeoQ7AS3obo2NC47Notyq5w%40mail.gmail.com.
Kent Perrier
To view this discussion on the web visit https://groups.google.com/d/msgid/nlug-talk/32768296.117329.1634661845701.JavaMail.zimbra%40mail.jobsoft.net.
Mark J. Bailey
Well, from a business-IT perspective, I don’t consider CentOS-Stream to be in the same support+stability category as RHEL/Legacy-CentOS. I see it more like a Fedora. What has always comforted me with choosing to deploy CentOS in enterprise-level, commercial environments was its 100% lock-step compatibility with RHEL (well, most of the time anyways), and this included CentOS’s RHEL LTS aspects.
With CentOS-Stream, that assurance (for me at least) no longer exists. Sure, I can have clients spring for RHEL. But a free 100% RHEL CentOS also enabled *millions* of VPS, and all sorts of other considerable efforts, to deploy CentOS at their core with genuine commercial confidence. But no more. This is why I’m totally rooting for Alma and Rocky, but they are not without some inherent risk exposure.
At the end of the day, we’re each at least entitled to our own opinion, right? :)
Kent Perrier
To view this discussion on the web visit https://groups.google.com/d/msgid/nlug-talk/918717943.117535.1634665211386.JavaMail.zimbra%40mail.jobsoft.net.
Tilghman Lesher
license of the various projects. Remember, you can modify software to
your heart's content, but as soon as you redistribute it, you have to
include the source code for those changes. There's far too much
contributed software for them to be able to switch to closed source.
As to who could take them to court over this, there are multiple
entities, such as the Free Software Foundation, as well as the
Software Freedom Conservancy, who have the resources to prevail in
court, not to mention the immediate revolt within their own ranks, for
trying to go Closed Source. There's just no way this scenario would
succeed.
--
Tilghman
Mark J. Bailey
Certainly, let’s hope not!!
But what concerns me is that, and just to reference the 2 you mention, each are 501(c)(3)s. Not sure about the latter, but according to the wiki for FSF, their total 2017 budget was $1,373,645. Now, granted, that is operating budget, and doesn’t reflect what might be their overall, current total legal slush funding. But in looking at just that annual budget, it pales in comparison to what IBM reported as $8.165B in cash-on-hand as of 06/30/2021.
IBM paid roughly $34B for Red Hat. It doesn’t take rocket science in the business world to expect that they might want to keep free alternatives to their mainline RHEL at some disadvantage, or easily vulnerable thereof.
All I’m saying is that push-come-to-shove, it might be a harder uphill battle than you think.
Mark J. Bailey
By the way, I never said anything about them taking anything closed source. Just making it a lot harder to remain 100% RHEL compatible (which is what really matters to me).
From: Tilghman Lesher
Sent: Tuesday, October 19, 2021 12:56 PM
To: NLUG
Tilghman Lesher
first place. Why exactly would IBM (and Red Hat) shoot itself in the
foot by trying to take RHEL closed source? It makes no sense
whatsoever, either in a legal sense, because they would be fighting
established copyright law, or in a PR sense, in terms of going back on
25 years of open source commitments.
Is there any justification for this line of thought? The fact that
they did something to CentOS isn't really an argument, here. They
were absolutely within their rights to alter what CentOS did, since
they own the property.
--
Tilghman
Tilghman Lesher
SRPMS are meeting their requirements under the license for the open
source projects. Under the various GPL-derived licenses, they are
REQUIRED to publish the changes for any distributed software. No
longer publishing those changes is, in effect, taking it
closed-source.
--
Tilghman
Mark J. Bailey
First off, Tilghman, I, like you, am entitled to my opinion, justified to you (or anyone else for that matter), or not!
Second, are you certain that the Red Hat *contributed* SRPM spec and related patch files are covered? Could they not just publish source patched tar balls of the core source components? What constitutes compliance here? And a good chuck of what the spec and patch files are also doing are not source modifying, but parameters, config files settings, and other items of convention (file/folder naming, locations, and what not), etc., which is a big piece of what the RPM approach sets out to do. Is that covered?
I think with CentOS, we’ve come to expect that the SRPMS and related components are covered, but that may have just been Red Hat’s convention to more easily comply with the GPL aspects. But that has largely gone away now, unless Centos Stream is including all that (and, honestly, I don’t know the answer to that, mainly because I haven’t had a chance to look).
I’ve worked with Linux non-stop since I first download the floppy images for the SLS distribution in mid-1992. That said, I’ve always thought that how Red Hat released and permitted projects like CentOS to build 100% compatible clones from their production line SRPMS was something living on barrowed time. And perhaps, Red Hat itself would hold to this position. But Red Hat ain’t really “Red Hat” any more, at least not totally. And I recall all too well the days of “Big Blue” IBM from late 70s up through the late 80s. Yes, they’ve played somewhat “nice” the past 30 years, but that doesn’t mean they will always continue doing so.
Maybe Centos-Stream will prove out for Rocky and Alma, and enable them to continue to mirror RHEL. I truly hope they can. Only time will tell.
Tilghman Lesher
compliance with the requirements of the license. But that would also
entail extra effort on their part. The SRPMS are easy for them to
publish, not only because it's their internal format for distributing
source, but it also allows contributors to contribute modifications in
that same format, which makes it easier (and cheaper) all around.
But could they expend a lot more effort to make it harder on everybody
else? Sure.
Again, I don't think that's particularly likely. It's really not in
their best interests to put forth more effort just to make it harder
on others in the ecosystem.
--
Tilghman
Mark J. Bailey
It will certainly be interesting to see how the ecosystem is affected by shift to Centos-Stream.
One thing I have noticed in the few times so that I’ve been on Rocky’s Discourse forum site is that what seems to be plaguing them most often are small, fine detail issues after updates to packages. If they are having build out of the Centos-Stream GitHub repo, as Kent was saying, then I think they are going to have a rougher time of it. Of course, and I must admit that I never followed the forums much on the CentOS package development side, but this may have also been the case with CentOS repackagings as well.
I need to follow up with my colleague down in Huntsville to see where the members on his team have gotten to with their Rocky evaluations. Since they are gov’t contracting, and apparently tapped CentOS quite a lot, they will undoubtedly have to be putting Rocky through the ropes.
Kent Perrier
To view this discussion on the web visit https://groups.google.com/d/msgid/nlug-talk/342962040.117950.1634675999022.JavaMail.zimbra%40mail.jobsoft.net.
Mark J. Bailey
No, Ken, I did not know that you were a Red Hat employee, but that’s good to know! :)
I had wondered about what resources Rocky might have backing it. I was aware that somewhere down the line, CentOS started benefitting heavily from Red Hat employee resources, even before Red Hat took them in directly. I can’t speak much to Alma as I really haven’t look at it that much, but I take it Rocky is working largely with volunteer resources?
And it is reassuring to see that Centos-Stream is at least including spec files. That I did not know. Does that include corresponding patch files want what not? And how will they mitigate, from what I understand, CentOS-Stream’s more frequent, “as ready” release cycles? It seems to me that that is where you can run into troubles.
Of course, nothing says you can’t hold off any updating until specific checkpoints in the release cycles (with respect the RHEL main branch). With CentOS legacy, I always felt that any updates were vetted well enough for RHEL main branch, so the risks were minimal. But perhaps that’s the value add for Rocky and Alma.
Tommy Kelly
--
--
You received this message because you are subscribed to the Google Groups "NLUG" group.
To post to this group, send email to nlug...@googlegroups.com
To unsubscribe from this group, send email to nlug-talk+...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/nlug-talk?hl=en
---
You received this message because you are subscribed to the Google Groups "NLUG" group.
To unsubscribe from this group and stop receiving emails from it, send an email to nlug-talk+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/nlug-talk/CALdmzXbbGuUest%2B18p2AqiRuV5azK-8f%3D%3D9xOb_pM8bdcS7OWQ%40mail.gmail.com.