Why does OEM versions of Windows even exist in the first place? How is a bloated, crapwared system ever going to be a better user experience than a clean install?
As for the why, have you ever tried installing vanilla Windows on a modern notebook? I know Linux has a reputation for bad driver support, but in my experience with Dell, Lenovo, and HP, a vanilla Windows installation will have no graphics driver, no network driver (neither wired nor wireless), sometimes the sound drivers are missing. It is a shitshow. Meanwhile, Linux on the same machine generally works out of the box.
But the answer to the question "Why do OEM versions of Windows even exist in the first place" is that without OEM customization, a Windows computer is a non-functional computer.
I think the driver situation has vastly improved from Windows 7 and even more since 8 and 10.
I have reinstalled Windows on quite a few laptops and desktops and have been quite surprised with how much works straight out of the box these days (mostly everything in my experience, bar some stuff like Fn keys). YMMV of course.
At the enterprise level it is quite common, specially since they want to have control over every little piece of software that goes into the stack, including updates.
I even worked at a few places that created their own UNIX flavors (mostly GNU/Linux based).
That's not accurate. We want similar things, but there is a large amount of minor variation in the type of problems they are using their computers to solve. Fortunately, the focus on portability and simple, well-defined interface allowed multiple goals to coexist. It's not perfect, but even when variations between UNIX flavors breaks something, it is far easier to track down file-location bugs and text interfaces.
The toxic type of political infighting often (but not always) involves cases where someone decides to skip these traditions; conflict happens when someone tries to "remove old cruft" or "unify the various old methods" without first considering who it will impact. Unfortunately, this has become a common problem in recent years when JWZ's "Cascade of Attention-Deficit Teenagers"[1] started to take over some projects.
But based on the slides and the date I assume the "same thing" all these UNIX(TM) vendors wanted was for their combination of proprietary hardware and software to win in the marketplace.
Gödel's Incompleteness Theorem might be abstractable to this. Any system so inaccessible as to be unhackable might be functionally impaired beyond usefulness.
> Also, I thought the consensus was that Apple was 'backdooring' iMessage by giving LE access to iCloud backups which include iMessage, not compromising iMessage directly?
Can you please expand on this? How would access to iCloud backup not compromise iMessage?
The backups store messages as plaintext within the [encrypted?] backup. This compromises the messages themselves, but not the iMessage protocol nor systems.
iCloud backups are not encrypted, and they include iMessage conversations. So the backups don't compromise the iMessage "service" or "application," but simply provide access to messages sent and received before the backup was made.
Can we please stop making Linux as a thing about choice[1] and stop fragmenting the community in the name of following a UNIX philosophy, the rule of KISS or some other nonsense? Are we really going to spend lots of energy the next n years discussing the best init system? Isn't this a pretty simple and solved problem by now?
Linux is just a kernel. It always has been and always will be. It is pretty useless without userspace. That's why we have distros. As long as multiple distros exist, they need some way or other to differentiate themselves. Differentiation in distros naturally presents us with choice. Linux has always been and will always be about choice. Any attempt to converge the userspace does not change that.
If you want an operating system where everything is developed in tandem, you might want to consider switching to one of the BSDs.
It seems to me there only needs to be one init system, otherwise it is a fragmented mess. Windows is gaining fast on Linux in the cloud/server space (SSH support, docker/container support, "Nano Server", more open source components).
Windows has no fragmentation issues in the core system.
Whether it is Systemd or System XVI im not really that bothered. Something needs to be decided fast however, and Systemd has a big head start.
I just hope it dosen't end up like the KDE/Gnome mess.
The mere existence of alternatives doesn't suddenly transform you from "unified utopia" to "fragmented dystopian hellhole".
Windows isn't gaining anything on the demographics that use GNU/Linux for cloud deployments. You're reading too many articles on HN and extrapolating that Windows must be curb stomping its competition because of some smart business decisions on part of Nadella and co. You're further making the assumption that Linux needs to compete with Windows. It doesn't.
It's not that there is fragmentation, so much as the problem space is quite open-ended and there are multiple solutions. Forcing a square peg into a round hole (systemd ueber alles) is a recipe for impedance mismatch and stagnation, not unification.
