Looking forward to 2025

The end of the year is always a good time to look forward to the next year — and so I will look at a few things I expect — or hope for — in 2025.

1. As IBM continues to subsume Red Hat into it’s corporate maw, I predict that Red Hat Enterprise Linux (RHEL) will stagnate and become isolated to IBM mainframe usage. Can you say Z-390? Red Hat has already closed most of its web content so that you need a paid subscription to access support and even the documentation. Also, having killed CentOS so that no truly free (of charge) Red Hat clone existed, there seems to be no real way forward for RHEL. Red Hat has always been the stodgy version of Linux and will get even more so under the IBM mantle. Regardless of any problems that follow, “No one ever got fired for buying IBM.”
2. AlmaLinux has taken the lead in Red Hat based distributions. It’s intended to be a rebranded version of RHEL, and it depends on the fact that all RHEL code is still freely available to clone like that. It does an excellent job of that and I expect it to take over a large portion of the RHEL market for users and organizations that want to try Red Hat without the expensive complications.
3. Fedora is the upstream distribution for RHEL, so it’s more advanced and has more fun and interesting features and applications. As a result, Fedora will be one of the most recommended distros for those users migrating from Windows and who want an advanced operating system without the stodgyness. I plan to stick with Fedora as it’s the best of the distros for my needs.
4. Intel will continue to lose ground to AMD in the CPU chip wars. Their current financial problems will weight them down and make it nearly impossible to recover no matter the size of the bailout. Intel still makes excellent chips so I expect them to be around for at least the rest of this decade. After that — well, who knows.
5. KDE Plasma continued to improve. I’ve used it quite a bit in 2024 but it still has a few rough edges. I continue to switch between Plasma and Xfce but am spending more time in the latter. I hope that the KDE Plasma developers will find a way to make it more responsive in 2025, It’s much slower than Xfce at rendering window moves, resizes, and context menus for applications like LibreOffice. In fact the waits are quite painful and waiting for the screen to redraw is so last Millennium. I’m just not sure how much Plasma can improve in the space of only a single year.
6. The battle for the GPU space will also continue. In an interesting turn, AMD says it won’t even try to compete in the ultra-high-end gaming market with Nvidia. Rather, it plans to compete in its favorite mid-range market. Intel also seems to have something new in the works, but as its perpetual micro-share of the market would seem to indicate, I don’t see it’s predicted GPU cards making any headway. I rather suspect that Intel will only succeed in improving its on-chip performance. The fact is, most of us need only very modest graphics in our computers. We just want to do our daily work and play simple games. The two groups that demand exceptional GPU performance are dedicated gamers who want that extra nanosecond of frame rate, and the crypto-currency miners who want that extra coin. Is that enough to support the market?
7. For all of you who dislike systemd (perhaps an understatement), there’s now a new contender for the Linux init system. The Shepherd is a new — if you count 21 years in development before a general release — “minimalist” init system that “herds” system services. Their words, not mine. All very cutesy. After reading the release announcement, some of the terminology reminds me of systemd. The most interesting feature of this old yet new init system is its ability to produce a graphical dependency map. Based on the announcement, even though this is a 1.0 release, there are still plenty of missing bits. Frankly I think the GNU project is, well, decades behind and will never catch up. Besides, I’m quite happy with systemd despite the “let’s not change anything” crowd.