This smells like IBM a lot.
In the comment section one Red hatter says the differences will be minimal for Alma and Rocky, though some embargoed fix might be out. Seems mostly ok (I’d say for most use cases even CentOS Stream is ok)
That’s a Red Hat employee that has nothing to do with code. The comments about emabargoed stuff appearing in Red Hat before CentOS Stream are for coordinated code releases to fix a bug that’s not been released yet. (e.g. there’s a remote code exploit in the network stack related to intel NICs. Intel will coordinate with people like Red Hat and MS to get the release out in a coordinated fashion, but the data Intel supplies is embargoed until the coordinated release.)
Rocky reports their release cycles are all tied to automation of the git repos that are going away. https://forums.rockylinux.org/t/has-red-hat-just-killed-rocky-linux/10378
So, while in theory someone who has a license can use source RPM’s to get at code, Rocky, and likely Alma, aren’t set up to deal with that as upstream sources. Plus, even though that matches the GPL (if someone sells you code, they have to supply source without restitrictions), I’d imagine the GPL doesn’t say that if someone sells a GPL’d product, they have to sell to you.
My guess is Alma and Rocky will figure a way around this, but id also guess it’s going to be tough.
Before we get too reactionary here, it could make sense to have people focus on the CentOS stream codebase for upstream dev, instead of Redhat having to manage upstreamed code targeting all the different releases of RHEL no?
And can’t Rocky Linux and Alma Linux just simply get source code via the partner program? Or does this change prevent them from doing so? You’d think that Redhat would want projects like Rocky and Alma around as a taste-testing lure for RHEL, considering that Redhat makes their money on support rather than RHEL itself.
RedHat already has no-cost RHEL licenses. The disadvantage is that it’s necessary to create a developer account, and one account only supports 16 devices.
Doesn’t this basically straight up kill distributions like Oracle Linux, Rocky Linux, Alma Linux etc.?
Which is just a good thing. Fragmentation has gone way too wide just to confuse the first-time users. Less projects with more working hands leads to a better solution.
The mobile linux is silly as well. 3 separate projects while none is ready. Still they all flood the aur with mobile apps.
Why there must be Cinnamon, XFCE and LxQt while they all looks 100% the same for end-user, but none supports Wayland, VRF or HDR? Those are standards which attracts first-time users than never-ending and confusing comparison between distros and DE’s.
I want to disagree with fragmentation being bad. It won’t take much to reduce the number of choices; Canonical and Ubuntu will surely be sold off to someone in 1-2 years. I would not want a few choices and then Microsoft just has to buy them and say “got you good! Bye bye Linux”.
