Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Rocky Linux: A CentOS replacement by the CentOS founder

github.com

Post reply on HN
Print view View original

531–540 of 555 posts

Re: Rocky Linux: A CentOS replacement by the CentOS founder

#531

Post by LinuxBender »

doublepg23 wrote: post #337
LinuxBender wrote: post #280

Earlier quoted context omitted.

RHEL and CentOS are both server and desktop. The desktop is just not flashy or shiny or bleeding edge. The default install is/was Gnome 3 shell (For RHEL7 and CentOS 7). I am not sure if this is still the case, but Redhat used to require 100% of its employees to use RHEL Workstation as the desktop.

I know a sales guy who had a choice of Windows or Mac, I think engineers use Linux.

That's good to hear. I felt kindof bad for the receptionist being forced to use RHEL. This was during RHEL3 and the desktop was not great.

Re: Rocky Linux: A CentOS replacement by the CentOS founder

#532

Post by acdha »

ghaff wrote: post #508
acdha wrote: post #379

Earlier quoted context omitted.

I was thinking more about _why_ people want that. If you're changing the system regularly, upgrading is valuable because you don't want to spend your time dealing with old software or backporting newer versions. Most of the scenarios where you do want that are long-term commercial operations where you need to deal with requirements for software which isn't provided by the distribution, and in those cases they likely…

>I was thinking more about _why_ people want that. I think there are two primary reasons. 1.) A developer wants to develop/test against an x.y release that only changes minimally (major bug and security fixes) for an extended period of time. 2.) The point release model where you can decide when/if to install upgrades is just "how we've always done things" and a lot of people just aren't comfortable with changing that…

Re: point 2, I don't know how different that is for stable distributions — e.g. if you're running Debian stable you're in control of upgrades and you can go years without installing anything other than security updates if you want.

Re: point 1, I'm definitely aware of that need but the only cases I see it are commercial settings where people have contractual obligations for either software they're shipping or for supported software they've licensed. In those cases, I question whether saving the equivalent of one billable hour per year is worth not being able to say “We test on exactly the same OS it runs on”.

Re: Rocky Linux: A CentOS replacement by the CentOS founder

#533

Post by davoneus »

jpalomaki wrote: post #136
mmrezaie wrote: post #3

I am really interested to know why should anyone go with this when Debian or Ubuntu LTS exist. The two later have not changed their policies in the last decade, and they have a clear path for upgrading. CentOS was always a clear choice for device drivers support, but I never understood the stability claims.

Not an expert, but I’ve understood CentOS was interesting for people who run RedHat for production, but want something free for non-prod hosts.

This.

We used CentOS as dev environments, and RHEL as production. It gave us the best of both worlds; an unsupported but compatible and stable dev environment we could bring up and throw away as much as we wanted _Without_ licensing BS. And when the devs were happy with it, the move of a project to RHEL was easy and uneventful.

And don't even get me started on the 'free' dev version of RHEL. It's a PITA to use, we've tried. It's also why we've halted our RH purchasing for the moment. Sure, it's caused our RHEL reps no end of consternation and stress but too bad. I've been honest with them, and told them that they are probably lying through their teeth (without knowing it) when they parrot the line that RH will have some magic answer for "expanded" and/or "reduced cost" Streams usage in "1st half of 21". That trust died when RH management axed CentOS8 like they did.

Re: Rocky Linux: A CentOS replacement by the CentOS founder

#534

Post by slipheen »

8draco8 wrote: post #515
jlgaddis wrote: post #424

Earlier quoted context omitted.

No kidding, right!? I mean, if you think about it, all that's really needed is for a "s/CentOS/Rocky/g" over all of the repositories. Then, the other 99% of the project is just waiting for the packages to rebuild and get sync'd on all of the mirrors that they could just will into existence with their minds. Really, though, let's be honest here: If they weren't spending so much time writing up press releases and comme…

But seriously, can somebody ELI5 this project. If Centos is just 1:1 RHEL with removed branding then what new bugs will show up in Centos that are result of that rebranding and will not be fixed by RHEL devs? Is there a code in RHEL that is also copy writed which had to replaced and maintained by Centos devs? What am I missing?

I'm not fully versed on all the things that would be needed, but at the bare minimum it would seem like you would need a bunch of automated processes for just the building - Import src packages, making sure that you copy in changes/patches when RHEL does. - Replace the RH trademarks in every package - Build every package and run verification tests for each arch - Build ISOs

You would also need I said if infrastructure servers that can scale to a large number of users for Yum/RPMs, etc.

Then you also need a set of servers for issue tracking and a way to break it out per package.

I wouldn't imagine that it is anything which can't be done, it just seems like there's a lot of little pieces that you would need to set up, and infrastructure you need to run.

Re: Rocky Linux: A CentOS replacement by the CentOS founder

#535

Post by effie »

curt15 wrote: post #339
effie wrote: post #283

Earlier quoted context omitted.

I'm wary of doing that, because in near future, Microsoft is likely to take over Canonical. You don't put all eggs into one basket. Always plan for escape, always have a plan B. Preferably one not relying on crystal-balling whims of a for-profit corporation. Rocky Linux, Alpine Linux, Debian, Gentoo, BSD, etc.

>Microsoft is likely to take over Canonical. This is not the first time I've seen this prediction. What is its basis?

There have been some interesting observations on HN and elsewhere, that Canonical for a long time didn't know what it is doing, starting and cancelling projects, but in last years it is lowering interest in desktop and is more focused on providing cloud server software and foisting its new products and methodologies on its users(snap). Some people see this as indication that Canonical is positioning itself to be bought for the best price. It makes sense, as Canonical has a large Linux user base but can't make money there. Microsoft is making inroads in Linux world is the most likely buyer.

Re: Rocky Linux: A CentOS replacement by the CentOS founder

#536

Post by quicklyfrozen »

Spivak wrote: post #408
quicklyfrozen wrote: post #364

Earlier quoted context omitted.

CentOS will be useless as a replacement for RHEL. Without the guarantee of binary compatibility, any CentOS Stream update may break your locally installed applications. And I only recall CentOS significantly trailing RHEL at the major version updates (e.g. 6 and 7). Other updates seem pretty timely, and the major version lag doesn't leave me vulnerable. I can see this being useful for developers who are building some…

I replied to you in another thread but nonetheless CentOS Stream isn't going to break your binary compatibility for the same reason that RHEL 7.3 doesn't break binary compatibility with RHEL 7.2. CentOS Stream is spiritually always the next minor release of RHEL. Unless you're the kind of person who pinned to a specific minor version of CentOS (which isn't the default and not supported for very long) you can use Cent…

They do specifically mention that some fixes may come to RHEL first.

I'm sure they'll try not to break binary compatibility, but as it appears to be somewhat experimental and targeted to developers, breaking updates may occur. Isn't that the point of this distro -- so such testing can take place before updates are rolled into RHEL?

So, fine for a developer workstation, but I don't see how it can be stable enough to use in production.

Re: Rocky Linux: A CentOS replacement by the CentOS founder

#537

Post by tinus_hn »

8draco8 wrote: post #513
znpy wrote: post #368

Earlier quoted context omitted.

in GIMP's defense, GIMP stands for GNU Image Manipulation Program, whereas names as Red Hat, Ubuntu and many others are not acronyms

Let's face it, idea for using word "gimp" probably came before they figured out what it stands for. Even now when everyone understands what that means and what consequences are behind using that word, there is still a lot of debate around that name and instead of changing it to something more marketable they are stubborn and are sticking to it. They even allow for bigger market fragmentation (Glimpse fork of GIMP) ju…

It was named after the Pulp Fiction character.

https://www.xach.com/gg/1997/1/profile/1/

Re: Rocky Linux: A CentOS replacement by the CentOS founder

#538

Post by matt-attack »

rfoo wrote: post #427
matt-attack wrote: post #418

Can someone explain why it’s not exceedingly simple to clone the existing Centos concept? Isn’t all the code that does the builds, artwork replacement, etc all open source? I would think that rebranding CentOS as Rocky is a rather trivial process of replatforming all the codebase and replacing any “Centos” with “Rocky”.

Because you need a lot engineers (so, hefty amount of money) to clone that concept? Basically CentOS concept is "freeze the version of all the packages and support them 10 years". But in real world people need new features, new bug-fixes, i.e. new versions. If you want to commit to that "never update" policy, you had to back-port everything people wants. In short, forking it is easy, but keep it attractive is not.

But isn't the backporting all done by RHEL engineers? I thought CentOS is just compiling RH's source RPMs.

Re: Rocky Linux: A CentOS replacement by the CentOS founder

#539

Post by matt-attack »

jlgaddis wrote: post #424
matt-attack wrote: post #418

Can someone explain why it’s not exceedingly simple to clone the existing Centos concept? Isn’t all the code that does the builds, artwork replacement, etc all open source? I would think that rebranding CentOS as Rocky is a rather trivial process of replatforming all the codebase and replacing any “Centos” with “Rocky”.

No kidding, right!? I mean, if you think about it, all that's really needed is for a "s/CentOS/Rocky/g" over all of the repositories. Then, the other 99% of the project is just waiting for the packages to rebuild and get sync'd on all of the mirrors that they could just will into existence with their minds. Really, though, let's be honest here: If they weren't spending so much time writing up press releases and comme…

While I enjoyed your sarcasm it doesn't help answer the question. CentOS did't do their own backporting of fixes, nor development. They take SRPMs from RHEL and do "s/RHEL/CentOS/" on them and then build the SRPMS into RPMs which get published.

Not saying all of the above is trivial but I'd think the code to do it literally exists and is itself open source.

Post reply on HN