The long-term strategy behind IBM's Red Hat purchase IBM's senior veep of software reiterated for Wall Street this week that OpenShift is the linchpin of Big Blue's overall multi-cloud strategy. Speaking at Morgan Stanley's Technology, Media and Telecom conference, Tom Rosamilia said the OpenShift container management family, developed by Red Hat that IBM bought in 2019, was key …
There are many things IBM has done wrong over its existence, but I think it's stretching it to blame them for failing to demonstrate in the 1990s running "1000s of copies" of an operating system that was little more than a curiosity when it first emerged at the end of 1991 at a time when only water-cooled IBM mainframes had more than 1GB of memory and the hardware was optimised for I/O throughput rather than heavy compute demands. The Internet consisted of dial-up connections for most of the decade, so any concept of "cloud" computing was somewhat premature.
There was some realisation at IBM of the capacity of the PC to subsume their lower-end products - hence their ultimately-futile effort on OS/2 - but if you're primary focus is on shifting iron, the idea of selling compute as a service is not one which will occur naturally.
Which is probably why the big three cloud vendors aren't hardware manufacturers.
This post has been deleted by its author
I like that term. "Anecdata."
I think it also has to do with the fact that one gravitates towards jobs that use familiar tools one can build on. I, for example, would not take a RedHat Linux environment contract because I have absolutely no familiarity with the environment, as I haven't used RedHat since 5.2 (yes, the 90s.)
The first half of the time after RedHat was pure personal use, and I had multiple machines, so I opted for "pure" Debian boxes, and Ubuntu was a natural follow-on to that when it came time to use what I'd learned commercially.
I on the other hand could not care less. Real men don't bother with binary package managers, they just use the source. And a lot of the source has RedHat and Fedora based people working on it.
Who do you think is ultimately responsible for creating (for want of a better word) systemd? Not some upstart ubuntu that's for certain.
The predecessor init system which Red Hat adopted for RHEL 6, creating the demand for the launchd-inspired systemd we all use today. This idea that Red Hat makes the most source is bunk, as you can clearly see how fewer packages comprise RHEL and how many of them are artificially crippled, needing replacements from third party repos. What is true though is that they have consistently contributed the most kernel code (from a single company) year on year for many years, and they do hire a lot of developers of core packages to keep them well-maintained.
However, Ubuntu also has good projects when it comes to server use. ufw is better for general use than firewalld, netplan is far better thought out than NetworkManager, AppArmor (Ubuntu-maintained now) is far better for creating targeted lockdowns of networked daemons than SELinux and their kernel live patching is better implemented. Also, don’t forget about ZFS being Canonical-supported, while Red Hat has practically given up on fancy CoW file systems, dumping btrfs in RHEL 8 in favour of the inferior XFS.
Don’t be too loyal to any one company!
Most Ubuntu users would not put up with an ancient kernel like this:
$ uname -sr
Linux 3.10.0-1160.42.2.el7.x86_64
$ cat /etc/redhat-release
Red Hat Enterprise Linux Server release 7.9 (Maipo)
You can get a more modern kernel for this platform, but that is very much frowned upon depending upon where it came from and what is done with it.
Another major problem is XFS, which is getting a formal deprecation because of this issue:
# grep xfs.*2038 /var/log/messages | head -1
Mar 7 04:55:30 kernel: xfs filesystem being remounted at / supports timestamps until 2038 (0x7fffffff)
https://lobste.rs/s/1wczzd/when_why_deprecate_filesystems
Mid-2023 IBM pull Red Hats support from LibreOffice. I suspect it is IBM that keep Apache OpenOffice alive due to it having their preferred type of license. What a pain this is, it is a decoy, a product without major updates since 2014. Perhaps they will put effort into Apache OpenOffice if they can make it online nicely, but it is so far behind, I suspect not.
"By rebasing our Cloud Paks on OpenShift, we've now moved all of our middleware to an environment where I can deploy on AWS, I can deploy it on Azure, I can deploy it on the IBM Cloud, and I can deploy it on prem,"
What he means is:
..but put it on IBM cloud. Go on. You know you want to. We'll sweeten the deal a little [inner monologue: until we get you on the hook].
I saw this so many times. Not just cloud - it's just the way IBM operate. I'd be interested to hear how many other readers have seen this recently.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2023
