Ah, those users. Users who want their OS to stay firmly in the background and let them get on with their jobs, and who aren't actually that bothered about "feature" upgrades every month or three.
Something is very wrong with a business when a big proportion of the customer base is viewed as a burden on the product.
The curse of products that are - in their foundations - over thirty years old and used by billions of people. To expand:
- any time Microsoft does fundamental changes to the look and feel or to management/configuration, enterprise users (the kind that buys tens of thousands of licenses at a time, and there's a lot of these) get pissed because on top of having to pay Microsoft for new software (to avoid not getting updates for EOL software), they have to spend significant sums re-training employees, testing upgrade paths, rework internal tooling...
- if MS breaks too much or forgets to add compatibility modes/quirks, said enterprise customers are going to be pretty pissed because that can sometimes mean they have to buy new machines costing many millions of dollars (the market for hardware sporting actual parallel ports or ISA slots is insane, because so much early 90s hardware runs on that)
- at the same time, desktop Linux distributions and macOS keep iterating and innovating without having that burden. Linux userland generally doesn't care a bit about backwards compatibility, and Apple announces BC breaks early enough / keeps support long enough that users can upgrade over time - aided by the fact that most hardware that people attach to Macs doesn't cost more than ten to fifty grands (=very expensive audio / video stuff) and has to be upgraded anyway to keep up with new features. That in turn puts pressure on Microsoft to keep up with innovation.
- on top of that, technical baggage keeps holding back Microsoft as well. Apple ditched Intel long ago and now reaps the reward - mostly, no transistor space wasted with long-superseded crap dating back to the 1980s and way better power management - while Microsoft is more or less at the whims of Intel's mismanagement and the utter crap that third-party ARM SoC vendors throw onto the market and dare call a CPU.
The obvious solution is to leave those enterprise customers behind on the legacy edition of Windows and launching something new. But Microsoft seems incapable of going to market with something new also. They created Windows 10x, and then went and merged the new parts back into the legacy platform.
The most natural time to have done this would've probably been going from XP → Vista/7, which broke a number of old things anyway.
For all that switching mainline Windows over to the NT kernel in 2k/XP improved Windows, the benefit wasn't as great as it could've been if they'd been able to unshackle the OS from backwards compatibility with all the preceding Windows/DOS stuff.
Microsoft had a good thing going, with the massive success of MS-DOS and Windows 95 it was imperative that the next generation consumer OS, Windows 2000, could run programs from their existing well-respected platform without users noticing the gap. (With OS X, you noticed the gap -- totally different UI style when in Classic Environment, and the Carbon API for easy porting was a mess that went unadopted by developers)
Really, they could have done a lot better-- a VM jukebox with everything from DOS 3.3 onwards available could have made the "most compatible Windows ever"-- even old software that assumed a 4.77MHz CPU clock could have been rigged to run natively.
Well yeah, and we still don't see Linux on every corner (android excluded). Linux desktop is a joke. You can either pick DE whick breaks compatibility every now and then (Gnome, KDE, ...) or which has more or less stopped developing (Xfce). I understand, it is nicer developing new things than maintaining and evolving old codebases, but I don't think MS needs to worry about Linux in this segment. And I'm saying that as someone who actually prefers any Linux flavor to Windows.
And it's not just UI where Linux keeps evolving. Just look at systemd - a fucking breeze to work with, whereas on macOS you're stuck working with plist's (annoying but OK), and on Windows it's literally the same crap from 30 years ago in the NT era. A Linux VM can do a cold start in well under 30 seconds, Windows needs anything from five to ten minutes and you can't optimize anything.
Windows services can be configured with dependency order and with delayed start. There are startup profilers.
I've seen and resolved slow boot times in misconfigured domains and with features such as roaming profiles. It's probably DNS ;-)
Sorry, forgot to mention I'm talking about Windows servers. Desktop indeed has gone quite fast.
That's a bit of a re-write of history. Mac OS tried to throw away all backwards compatibly on day one but major software vendors told them to pound sand so they had to spend another year building out Carbon. Not to mention that Mac OS X itself is built on NextSTEP.
Linux breaks front-end compatibility so much that never seem to make much progress on stability.
> on top of that, technical baggage keeps holding back Microsoft as well. Apple ditched Intel long ago and now reaps the reward
It's not technical baggage that is the problem. It's strategic issue. Microsoft had a fully functional build of Windows 7 for ARM a long time ago. The problem is they never just released an ARM version of their OS and let developers port their code to it. It was always paired with a new API, the windows store, a locked down version of the OS, etc, etc. This might be specifically to avoid pissing off their hardware partners.
Windows NT was always multi-platform and they always made sure it would build and run on non-Intel CPUs. It has never been a technical issue.
New features or APIs in MacOS are adopted much faster than in Linux world, too (see also: Wayland. Or, on similar complexity level, global dark mode).
My most recent experience with this was updating the Raspberry Pi OS -- the latest version completely breaks RetroPi and I had to reinstall and revert.
Yes, and other than that the only non-workaroundable - but reasonable - break has been the removal of 32-bit x86 support. The PPC->Intel and Intel->ARM transitions had fully fledged and (at least in the latter case, didn't have a Mac at PPC time) high performant and fully functional runtime emulation/translation to ease it.
> Microsoft had a fully functional build of Windows 7 for ARM a long time ago.
That depends on your definition of "fully functional". Apple invests insane amounts of effort to tape out good hardware that isn't riddled with weird bugs and edge cases. They don't even publish much documentation, and yet not even a dozen people (iirc, please correct me if wrong) managed to reverse engineer enough to get support merged to Linux mainline.
In contrast, the utter bullshit that the rest of the ARM world dares call a CPU is so riddled with bugs that it's a nightmare even for the vendors to keep their custom kernel patches / BSPs up to date, and completely forget about ever upstreaming it to the Linux kernel. And now imagine someone has to write Windows drivers for all that buggy crap.
The completely sorry state of how the entire "embedded" world works is holding everyone back so immensely that it takes companies the size of Apple and Amazon to show that the problem isn't ARM itself - and unfortunately, Apple doesn't sell to anyone else and Amazon only makes server CPUs, which means there is zero market pressure, neither for a re-imagine of x86/64 without legacy crap in the instruction set and die, nor for anyone with an ARM license to tape out a SoC at above "barely works, fix the rest in software" quality and decent open source Linux and Windows drivers, nor is there much pressure on AMD/Intel to improve because as long as there is no quality competition from ARM they don't have to fear anyone else eating their lunch.
Sorry for the rant, this topic has annoyed me ever since I tried putting an at-the-time modern kernel to an at-the-time-current Mediatek based phone and miserably and utterly failed a decade-ish ago. And nowhere I have looked at in the embedded space since then has shown me any sign of improvement. To the contrary, even projects as popular as Raspberry Pi couldn't arse Broadcom enough to check if the PCIe implementation passed standards validation, they only cared if the USB (and iirc, wifi too) chipset worked. J F C
I don't know how modders manage to get Windows 10 working on Android phones without any source code yet they still somehow manage to get it running with various degrees of success:
https://www.xda-developers.com/renegade-project-boot-arm64-w...
And yet, I don't disagree with your rant. The entire "embedded" world is a bit of a mess when it comes to hardware and software. Apple can only do what they do because they manage most of it themselves. Now Microsoft can, and has, also done that to various degrees of success but they're still both hamstrung by their own strategic decisions and by the problems you've described because they're still outsourcing much more than Apple is.
Evidently not, considering there were 4 such releases. The earliest was on July 29, 2015, and the latest was on November 16, 2021 and will be supported until January 13, 2032.
Curious what they will do about Microsoft Store apps in such a release. It's becoming more of a pressing concern. Maybe Windows 11 LTSC will have a "filtered flathub" situation similar to Red Hat / Fedora?
They really didn't.
Full disclosure: I installed Windows 10 Enterprise LTSC 2021 on a brand new laptop earlier today :)
For this specific device, my wish is a Windows-compatible OS that's going to shut up and stay out of the way, and it seems the LTSC variant of Win10 is the closest Microsoft are willing to provide at the moment.
That LTSC doesn't come with Edge, Cortana or the Windows Store, well that's just the icing on the cake.
Looking back at the comment, yes, you are right. Both lines could share under the hood stuff, the NT/2000 line with clean UI and the ME line the colorized cartoon one. Or, they could split the line with Home and Pro, both of them they still have.
Yes - I’ve been wondering since win10 why Microsoft thinks the enterprise edition needs an xbox app preinstalled.
Windows should not be a Mac OS clone. But every single UI update sends it in that direction.
It was marketed as a "Zero New Features" update [0]. Only performance, reliability and back-end upgrades (and it was a packed release for sure that completed the 64 bit transition).
EDIT: Anyone who agrees, upvote this comment! Who knows maybe someone in Redmond will see it! You never know!
[0] https://9to5mac.com/2022/02/26/apple-work-looking-back-at-wh...
Well, they capitalized on their market share immencely with Azure. At the beginning it was pretty ****, but a lot of customers signed up immediately. Not because that **** is better, but because Microsoft already held all their eggs.
Windows is the linchpin in that strategy so I don't understand why they've moved organizationally moved it so far away. It's relegated to being Bing adjacent.
https://gs.statcounter.com/os-market-share/desktop/worldwide
In case MS marketing reads this thread: you can simply fix the outstanding, sore-thumb problems with your product in its own way and gain both types of customers. Instead of poorly mimicking something they never wanted in the first place.