Funny...
Because I have never heard of a continent trying to force me to create a "continent-account" before...
Also continents don't suddenly decide to 'upgrade' and force you to buy a whole new tectonic plate...
Say what you like about its role in the destruction of civilization, the net is still good for a few party games. Take bets on when the "Wintel Empire" was first reported as under attack, and by what. Then go and find out. For younger readers, the Wintel Empire bestrode the IT globe from the mid 1980s to the 2010s. It still …
The separation into tectonic plates encouraged the differentiation of accounts making it very hard for anyone to travel unaccounted.
As an example, trying to enter the Americas without an account can be, well, problematic and leads to deportation. But even when you have an account you may not be able to login, which leads to the same problems and deported situation.
I've the opposite problem. Due to various responsibilities, jobs, and avocations, I have multiple accounts that don't all align in time and space. It is literally impossible for me to reconcile the peculiarities of this M$-forced paradigm. So I wander about the interweb net on various different devices, with various different accounts, doing various incompatible things. There are some days I don't know if I'm a bit or a byte, I can tell you. The nonsense is stultifying.
For younger readers, the Wintel Empire bestrode the IT globe from the mid 1980s to the 2010s.
<pedant>
I think you could reasonably argue it started with Windows 3.0 in 1990.
</pedant>
Even though MS were dominent, there was much more competition in the 80s, DR DOS was still a thing, AMD had a decent line of competetive performing clones. You could tell the way the wind was blowing but it didn't really get there until the early 90s.
Wishful thinking. Most people on a PC still have Windows, some MacOS, a small minority Linux. Despite Nadella's attempts to kill Windows and have everything in Azure.
Smartphones doesn't matter. People there don't choose an OS, they choose the shiniest model they can afford. After all, the form factor allows limited interaction, and they are used by the mobe, more than they believe they are using it.
I beg to differ. My first smartphone was Windows Mobile and I loved the simplicity of its UI. Then I had various Androids and I was ready to throw them out the nearest window because of the inconsistencies and gaps in how it worked. I’m now a committed iPhone user (albeit the cheapest model) because it’s all integrated and just works.
I love the continuous lie that the iPhone "just works".
It only means you've gotten used to the stupid idiosyncrasies of iOS, versus the stupid inconsistencies of Android.
Take your blinders off.
Don't go telling me that, from an Messaging screen, when you attempt to discover your saved contact's details that the *constant* left-right-left taps of acknowledgement and app closure "Just works"tm!.
If so, you're just lying to yourself. Seek therapy.
I thought it had a micro-kernel with BSD userland.
Micro-kernel is structure of how userland talks to kernel.
Unix is a system.
As noted, as long as your O/S compiles and complies with the (paid for) Unix tests, then it's a Unix O/S.
BTW - The BSD's are Unix, they just didn't pay to get the Unix © ® ™ marks.
Apple took Mach and made many changes to it, including turning it into a hybrid kernel to make XNU.
Apple also copied a bunch of BSD software and parts of NeXTSTEP and many other programs under weak licenses to make Darwin, making their own modifications - with the result being clearly not a BSD.
Yes, a Mac actually IS Unix. Why do I know? I was involved in the certification of MacOS being Unix. You need to run a WHOLE bunch of tests to make sure it "conforms" to the specification that Xopen promulgates. The tests check EVERYTHING in minute triviality just to make sure. This included errors. It is actually quite rigorous.
Yes, Xopen holds the UNIX trademark, and is quite particular in who is allowed to use it.
Pro Tip: If you want to specify a MacOS in a purchasing document, but don't want to say MacOS, just say it needs to be UNIX at a price, and there you go! All the other alternatives are higher priced than a MacOS machine.
Yeah, about that "Mac is Unix" line. There's some argumentation about how much of the testing the current versions of MacOS conform to, which is apparently a lot less than when the certification was first done. Word I've read is that it can only technically claim for that to still be true, but it is a nice line to have in the product info.
Edit: On a quick check, I may well have been underselling what that cert is still worth. Eh... https://www.theregister.com/2024/10/11/macos_15_is_unix/
If it comes to that, Windows would probably pass the Unix test these days (or get close), specifically with WSL 1.0 and an appropriate distro loaded on top. No VM, just a Unix user land (alongside a Windows user land too). That might be an unconventional view of what a Unix is, but AFAIK the standards don't prohibit non-Unix userlands being on the system alongside the Unix stuff.
Yes, I know Linux isn't quite up to standard, but then what is?
Windows SFU was there specifically so that Windows could qualify for US Department of Defence contracts. I don't remember it being especially good or frequently used. WSL is heaps better.
The DoD were the branch of gov that mandated open standards for hardware and software, inspiring POSIX and the adoption of VME. That did us all a huge favour.
Android happens to use the kernel, Linux, but Android doesn't really allow access to its /sys, /dev, /proc or SYSCALL interfaces unless you have root - most software can only access whatever the Java Virtual Machine and/or the "native system libraries" allows them to access (the result is that whatever kernel schedules the processes etc is mostly irrelevant - google might swap to Zircon if it becomes feasible to do so and if they do, nobody will notice much difference when it comes to purely Java software).
What Android is missing is GNU - although it is possible to install GNU bash and many GNU packages via Termux - the very design of Android prevents full GNU/Freedom from being achieved.
MacOS is a Unix, as apple has paid the "open group" for the Unix certification - which consists of making a payment and also passing some ultra-proprietary POSIX test suite.
When I first came across Arm chips 15 years back, I realised it was the future. One day all laptops will run on Arm, I said. Apple have done that (and by golly it’s fast and battery life is something my works HP can only dream of). There’s no reason (apart from pride) that Intel couldn’t license Arm cores and build chips themselves.
The problem for moving Wintel from x86 to ARM is down to Windows and backwards compatibility. Windows has 30ish years of legacy software that it needs to remain compatible with. Jumping to ARM would lose all of that, or at least make it much harder and require emulation.
I wish MS would make a clear break though. Windows desperately needs a modern, stripped down version built for speed and performance. Unfortunately, their focus is on data mining and AI.
Apple are not perfect (there’s a phrase doing a lot of heavy lifting), but they managed to go from power to intel and intel to arm whilst still allowing backwards compatibility. In the PC world some low spec stuff may need to be certified to only run the new and not the compatibility layer as entry level pricing and capacity for windows is lower than Mac, but clearly if apple have done it twice there is a path that can work. Radical change in a sensible direction is something I stopped expecting from politics or the Windows bit of Microsoft when I was in High school and I am yet to have this assumption challenged, I would be very happy for this to change but I will not hold my breath.
And Windows on ARM has X86 emulation just like Apple has Rosetta. If you're going to give Apple credit for that, you have to give it to Windows as well. You're also giving Apple too much credit for their emulation. They had it in some versions of their OS, then they removed it. If you wanted to run some PowerPC software five years after the last PowerPC model was released, not that crazy if you kept yours around for that time, sorry, that compatibility was removed in 10.7 and you can't use it anymore. Apple relied on most users replacing their software before then. They're likewise planning to remove their X64 emulation in 2027.
This is a bigger challenge for Windows because there are many people who use it because it can still run some piece of ancient software. While I know some people who have the same situation with Macs, the only person I know who is still doing it has a collection of PowerPC Macs he still uses because newer things won't run the software. Specifically, he works in audio production and became attached to specific pieces of software. He could probably use something newer but has decided that he'd rather stockpile PowerMacs than learn two decades of new interfaces.
Granted. Apple can take deserved credit for testing theirs thoroughly and not releasing anything partial. Microsoft can be blamed for releasing the version that only worked on some things and improving it by stages. They still have it, though, which a few posts here either are unaware of or deliberately ignoring.
I know this is late but what I was referring to is not simply the technical capability to run old apps but the fact that apple took most users with them (admittedly a smaller group than windows users). Not only that other vendors of cross platform and apple targeted software also followed. Microsoft seem able often to deliver the technical capabilities but often release too early with too many bugs or fail to bring hardware or software vendors with them or fail to educate the user base or all of these. This is harder for an organisation that has more variation and does not control the hardware but developing an OS for that ecosystems is the purpose of the Windows unit at Microsoft and they have considerable resources at their disposal. Windows 11 runs on ARM I have run it on am M1 Mac and a raspberry pi but I cant easily buy a cherap arm windows 11 laptop I would want to own and there are no competitively priced arm windows desktops or laptops that can match apple battery life and performance in one package and only a tiny handful that can do one or the other. In that market Windows stays on x86 as its not worth the effort if you arent get the performance or battery life bump or the price drop.
That Is why this is frustrating Microsoft have the capability to do this qualcom, Intel and TSMC have the capability to do this. They aren't doing it.
> Apple are not perfect (there’s a phrase doing a lot of heavy lifting), but they managed to go from power to intel and intel to arm whilst still allowing backwards compatibility
Actually, Apple managed to go from M68k to PowerPC, from PowerPC to intel and from intel to ARM, all whilst still allowing backwards compatibility during each transition.
Indeed. As a long time user (and for a while, running a dealership) I went through all those phases. Not to mention, the sub phase where the 68040 introduced separate data and program memory spaces which caused some interesting breakages in self-modifying code (as in, Apple had told developers "don't write code that alters itself, it'll break at some point", and developers did just that and their software broke at some point) but at least the chip had a backwards compatibility mode you could turn on and off (needed a reboot, and it made things slower, but it allowed you to run broken software until the fixed version came out).
The 68k to PowerPC phase involved custom support in the silicon - they were one of the three companies in the group that created the PowerPC. And they made it easy to provide "fat binaries" containing both PowerPC and x86 code instead of having to ship multiple versions.
Then PowerPC to x86 involved software - which was written by a spinoff from Manchester University, I applied for a sysadmin job with them back in the days.
Ah, another stroll down memory lane.
Even if they have custom native modules .NET Framework 1.1 applications can still be made to work on .NET Framework 3.5, you just perform an administrative install and ignore all the prerequisite prompts. It is also still expected to be maintained until at least 2029 and most of it will also work on Wine too, if installed correctly. Even if Microsoft abandoned it tomorrow, .NET Framework could also be replaced with Mono to keep things working longer.
Unless you mean .NET Core, in which case, you’ll need to mess about a bit more.
Windows on ARM runs Windows/Intel binaries just fine. They've done the same thing Apple did, and done a good job of it.
They've also made it pretty easy for software to build for Windows/ARM. Their dev tools are perfectly capable of producing binaries for both platforms.
I'm not entirely sure what you mean by a modern, stipped down version. Windows 11 is no slouch.
Last month I finally upgraded from Win10 to Win11. The Win10 install was at least two years old, so plenty of time to pick up cruft.
I clean installed Win11 and it runs noticeably slower on the same hardware. That's after turning everything off that I could find. General performance is worse, gaming performance is worse. That's compounded further when I compare it to gaming in Linux, again on the same hardware. Technically worse hardware as Windows is on an NVMe drive and Linux is on a SATA2 SSD. Certain games run better on Linux. Whether that's due to optimisations or just that Linux uses less resources itself is up for debate.
Windows 11 may not be *bad* as such, but it's not *good*. It's adequate at best.
Does nobody here remember FX!32, the JITC emulator that DEC released in '96 for executing x86 binaries on DEC Alpha processor boxes running NT? When I worked for Microsoft, I had a DEC Personal Workstation 500a ("Miata") running NT4 that had no problem running the x86 version of Quake under FX!32. And in the pre-release version of NT5/W2K for Alpha, FX!32 came bundled with the OS. It all worked seamlessly.
/that box went on to run FreeBSD for Alpha
In the past they did. They sold most but not all to Marvell and a lot of it came from DEC (maybe all?). The DEC StrongARM was an attempt to do a low power alternative to DEC Alpha (which had NT and even a beta NT 64 bit) in 1996, so almost 30 years ago.
Intel may even still have some sort of ARM licence.
Only 15 years back? Intel and others (486) were losing platforms and products to ARM nearly 25 years ago. The Apple Newton about 27 years ago. The Acorn Archimedes desktop was nearly 40 years ago (1987) and had RiscOS and UNIX.
Apple was late to Smartphones in 2007 which were almost all ARM, So it ran ARM with iOS derived from MacOS. Over 20 years ago.
But unlike Apple, Windows on ARM has been a failure. Windows NT though starting as multiplatform became locked to x86 partly due to the purveyors of business applications, who mostly still don't support MacOS or Linux. WinCE on ARM was relatively successful as no-one expected to run desktop apps on PDA etc.
MacOS on Arm has been enormously helped by the Rosetta compatibility tool and the Universal packaging that allows apps to code for Intel or Arm or both. There has been little fuss as far as users can see. I’m not aware if Windows has a similar strategy.
The advantage Apple has is that it controls the entire stack from processor to OS (and a lot of the applications). This means as a company theu could driver the changes, rather than relying on others. So they could implement the abstraction layer that allowed legacy code to be run on a new processor architecture with minimal performance loss. Also most MAC products were remarkable homegeneous in terms of hardware
microsoft coulkd do something similar. they couldlicence, build and control their own ARM architecture. The problem is for them, for many years windows has been licenced to myriad hardware manufacturers, meaning it has to have all sort of fixes to support the different variants. there is no straightforward wintel architecture. This makes porting to another platform more difficult since you can only support a small subset of the existing variants totally, and others vonly partially
In many ways Window is a victim of its own success, which was basically sell it to anyone. Even small hardware dependencies such as the windows 11 TPM support creates huge ripples in the support cases. MAC on the other hand deliberately went the other way, of keeping everything under control. This meant in terms of market share they lost out, but in the longer term they have far fewer issues porting existing machines and applications
Unix-style OSes have also been ported to many platforms. It's not so much a matter of having control over the full stack, it's a matter of having a stack designed with distinct layers and well defined interfaces.
As a consequence I can run the same distro on an Intel laptop of a Pi and the same apt install command will install the Intel or ARM version of the application and its dependencies as appropriate, I'm not sure Windows could manage that.
I never said there was binary compatibility. Intal and ARM do not use the same instruction set and hence their native binaries are different.
If you have "binary" compatibility being marketed then either you have a dual weight "binary" that eats your download and disk and the OS picks out the correct half to run or you have an emulation layer that eats some of your CPU on one of the architectures. Done properly there's no problem. You install the correct flavour of OS and the rest works without wasting resources on your computer.
> I never said there was binary compatibility.
At the time binary compatibility of platforms was given as a strength of the Wintel platform in the Windows vs Unix battle.
> dual weight "binary"
Not sure about these, multi-(Wintel) platform installers have been a thing for sometime with some shipping a single installer that installed either a 32 or a 64 bit binary depending upon the installed Windows version. I would assume MS may do similar in future having a single installer for x64 and Arm binaries.
Poor binary compatibility (as in slow) was what sunk Itanium.
As well as long term stable binary compatibility, Windows did well to make sure installers continued to work. Once they sorted out the DLL hell of the early days (when we all realised that shared DLL's were pointless and hard), installing and uninstalling Windows software has worked across multiple versions of the OS.
Whereas Linux has been beset by too many different and incompatible ways of installing software, and it's not been helped by the chaos in build (autotools, which is truly ghastly, cmake which is moderately ghastly, etc). Building software should be easy, but it's not a reliable way of distributing software to every user, and binary distribution is a messy task too made a whole lot worse by glibc's policies of breaking old binaries. For a kernel with a famously stable system call interface it's unbelievably hard to take a binary from one OS running the kernel and have that run on another OS running the same kernel (the kernel isn't the problem, it's everything else).
Microsoft still haven't worked out self-inflicted DLL hell. Their non-solution to the .net mess that evolved out of the ActiveX/COM mess which evolved out of the DLL mess is still... a total mess, just much worse than ever. The WinSXS directory and it's hundreds of thousands of directories and copies of the same and different version of DLL files (.net assemblies) and nothing has been solved whatsoever, just pushed into an unmaintainable mess that when it goes wrong will prevent arbitrary parts of the system or applications from working as expected.
To be fair to Microsoft, if you look at the various GUIs many Linux distros now use for app installs, their equivalent would be the Windows Store, which is absolutely capable of hosting and selecting the correct version of an application and it's dependencies (albeit the 'Windows' way, which is more equivalent to Flatpak, but worse in almost every regard).
The problem is, (almost!) nobody does that - the standard methodology for Windows since before it was even a thing has always been to install and/or run manually obtained binaries from any old source - whether that be a floppy disk, optical media, flash drive or the web. Try doing that on Linux, and you'll have similar issues a lot of the time.
The difference really boils down to the fact that both the people who make Linux apps, and the people who use them, know and prefer to use better ways to handle software installation and management, and meanwhile, Windows is stuck supporting the free-for-all that developed in their playground, with no ability to do much about it unless they lock things down to their preferred method, which isn't going to play nice with the 'backwards compatibility' mantra here. They've tried this with 'S mode' in some of their first-party devices, but all it did was cause a spike in people googling 'how to turn off Windows S mode' the first time they wanted to install some random app they got elsewhere, and the impression I get is that Microsoft are edging away from it the way they do when they plan on ditching it in the medium-term.
I know that the likely response from Linux users (of which I am one) here would be to point out that on Linux, apt, packman, flatpak, Snap Store, Gnome Software or whatever CLI or GUI tool you use isn't limited to just a single source, but is capable of pointing to multiple repositories, public or private, while the Windows store is more akin to the Apple App Store or Google Play, which is a fair point. However, looking at it from Microsoft's perspective, I can certainly see why they'd want to retain control over exactly what software can be distributed via a 'store' under their brand. Given they aren't solely distributing open-source (or even at least 'source-available') software, they can't rely on the 'many eyes' approach either. On the positive side, so far as big tech app stores go, Microsoft's are some of the best deals out there - just a 15% cut of sales made via store infrastructure.
Microsoft could have very easily built into the "Windows Store" interface the capability to pull software from different sources but all to be managed and installed in a cohesive manner. This would have greatly benefitted large organisations who could run their own store supplier and even independent software vendors who could provide their own software through their own store linked through the "Windows Store". However, as usual with Microsoft, greed kicked in and they wanted everything to be their own systems and nothing else, and to slowly turn everything into a regular licensing income system to the exclusion of everything else.
"As a consequence I can run the same distro on an Intel laptop of a Pi and the same apt install command will install the Intel or ARM version of the application and its dependencies as appropriate, I'm not sure Windows could manage that."
Which part don't you think Windows could manage? Picking the processor architecture you chose from the packages it already built? It definitely can do that. Anyone can do that, it's one switch statement. Microsoft's got their command-line tool retriever, WinGet. That does that. But you probably meant that the running the same distro part is the impressive bit, in which case you're right about that and wrong about the basis.
You can run the same distro on a PC and a Pi for one reason only: the Raspberry Pi is extremely popular so distro creators go to large efforts to make sure it has compatibility. If that means distributions get held back to fix something Pi-specific, that is justified because there are so many Pis around. Try running a distro of your choice on some other ARM board and you'll have a different experience because compatibility is actually hard. That's why there are three categories of ARM SBCs out there: mainline boards where you should at least be able to patch things in if you have the knowledge, Raspberry Pi which can do whatever it wants because it's popular enough that people will work around it, and the rest of the boards where you have to be much more careful. Windows on ARM has a similar problem which is one reason they're only running on Qualcomm SoCs so far, although I have to give Microsoft credit that they're only working with standard firmware, which also means that one of the types of machines where Linux on ARM is easier to get started is those devices already built for Windows on ARM.
>MacOS on Arm has been enormously helped by the Rosetta compatibility tool and the Universal packaging that allows apps to code for Intel or Arm or both. There has been little fuss as far as users can see. I’m not aware if Windows has a similar strategy.
Yep, MS has an equivalent of Rosetta, works pretty well.
As far as I can tell, Apple's real advantage in Arm has come from the fact that they design their own licensed Arm cores while Microsoft is trying to do the Android-style thing and sell Windows on Arm to manufacturers. So whenever some manufacturer wants to sell a Windows on Arm machine they need to make sure it runs on every part of their system or buy an existing/license a reference system design and repackage it. It's far more expensive to design their own cores and complete systems (and get Windows working 100%) so only the biggest chip players have tried (like Qualcomm's new Snapdragon laptops) and the off-the-shelf designs are apparently not very fast and don't get updated for years.
So Apple's vertical integration and long experience with iOS makes their shift to Arm comparatively simple and allows them to optimize and improve things incredibly quickly while the Windows world just isn't iterating anywhere near as fast because existing x64 laptops are Good Enough.
"Windows NT though starting as multiplatform became locked to x86 partly due to the purveyors of business applications"
In part, yes, but mostly that has happened as a result of Redmond's own restrictive policies. For decades MS has done all it could to lock software developers and vendors into the Wintel eco system. Now they have painted themselves pretty much into the same corner.
Then there's the inertia of the market. The saying that "Nobody ever got fired for buying IBM" is now "Nobody ever got fired for buying Microsoft. It's a boardroom-safe option.
I am not sure this is correct.
How have Microsoft locked software developers into the Wintel ecosystem? There is nothing like the restrictive polices that the are around the iOS App Store. Okay, things have changed more recently but in the context of this article anyone and their pet dog could develop an application to run on Windows (or MSDOS) and not require any approval or certification. They just wrote it and it became freeware or a trial on a CD stuck to a magazine or you bought from a shop.
This is one of the huge benefits and why Windows has grown the way it has. The biggest win came when Microsoft managed to get the OS pre-installed and then the dodgy deal to force it.
That is nothing to do with developers. Both users and develops then grew as the platform became the most widely adopted.
To be clear on this I was selling computers with the BBC Micro, Acorn Electron, ZX Spectrum at the time. Then can the first mainstream "PC", the Amstrad PCW256. Ridicule it as much as you want but it (mostly) worked and users bought a box and plugged it in. There was quite a range of applications available, the disks could be easily handled and it came with a printer.
This was followed by the PC1512 with Gem Desktop. This was very quickly eclipsed as Microsoft and the "Wintel" machine took over mainstream computing.
As the user base grew so did the numbers of developers and applications.
Apple was late to Smartphones in 2007 which were almost all ARM, So it ran ARM with iOS derived from MacOS. Over 20 years ago.
You are forgetting that Apple basically saved ARM, becoming an equal shareholder with Acorn with VLSI as a smaller holder. I believe the original split was 40/40/20 or somesuch. Basically, Apple was committed to ARM and I don't think they seriously considered any other processor for the iPhone, even after the non-success of the Newton.
And basic mobile phones back then were not all ARM; quite a lot of them seemed to be using x86 type cores. MCS86 or 96 was it called? Even the "smart" phones (I had an Orange branded thing) didn't all use ARM by any means.
M.
>Windows on ARM has been a failure.
Well, Windows 11 on ARM seems pretty well sorted. But yes, historically, MS hasn't really realised the potential of Windows on ARM.
When MS showed off Windows 7 running Office and printing to an Epson printer nearly 20 years ago, many of us then thought that this was the beginning of MS pushing desktop Windows on to ARM. Except, they didn't follow through. I think MS bought an architecture license from ARM, which meant they could have done what Apple has done more recently and built their own range of ARM-compatible silicon. Instead, they didn't and we got WinRT on tablets. Bit of a let down really.
I guess there was no real pressure to do so because at the time Intel were still good, and there was little else out there to compare it to. It also wasn't that long after Apple had switched over to Intel either (we were mourning the passing of Mac/PPC).
Windows on ARM today works. Unless you go looking, you can't tell. Intel binaries work fine, the only blemish I've spotted is tha the Google Earth Pro desktop application works but doesn't look good on the ARM Surfacebook I've tried it on.
But unlike Apple, Windows on ARM has been a failure.
Has it?
There's a few Windows on ARM laptops out there, and until last year Qualcomm had an exclusivity deal (but not that good silicon ironically), but that's since expired, so I wouldn't write it off just yet.
There’s no reason (apart from pride) that Intel couldn’t license Arm cores and build chips themselves
Intel has long held an ARM architectural license. They even owned the fastest ARM cores (StrongARM) 20 years ago - that's what Steve Jobs wanted for the first iPhone, but Intel would only sell them x86 so Jobs told them to pound sand. Intel might have followed a VERY different path had they made that deal.
But the idea that Intel can just "build" ARM chips as if that's a solution to anything is highly flawed. Yes you can run Windows on ARM, but consumers have demonstrated they don't want that. Despite Qualcomm's ARM PCs being better in some respects than competing Intel/AMD laptops, people don't want to deal with performance hits and compatibility issues when running x86 code - and games mostly won't work at all due to stuff like copy protection and anticheat mechanisms built into many of them. Qualcomm PCs have a tiny niche of the Windows market.
Apple was able to manage the migration because it was a migration - the x86 Macs were going away, so if you wanted to stick with Apple you had to follow them to ARM. It helped a LOT that the ARM Macs were better in almost every way than the Intel Macs, of course, so there was little arm twisting involved. Microsoft isn't migrating, it just added a second platform. Something they've done before and they've always dropped, so people are right to not fully trust that Windows on ARM will be around five years from now.
The integrated graphics on M1 Max (let alone M4 Max) are significantly faster than the integrated graphics on any x86 CPU. They are close but not quite as fast as the fastest discrete laptop GPU. They are no match for the fastest discrete desktop GPU, so it depends on what level of "3D graphics games" you're talking about.
Of course, but the emulation layer would then also have to provide a transparent and bug-free layer for 3D acceleration, and that's not often the case. Not sure if Apple has done this well. as I haven't tried any serious games on the M1 yet. I know that platform-agnosticism is coming along though, so this may well work better nowadays. (You have the Steam systems, for example.)
"There’s no reason (apart from pride) that Intel couldn’t license Arm cores and build chips themselves."
They did: StrongARM, acquired from DEC, and later XScale. Then they decided focusing on x86 was the way forward, with the intention being to extend its reach to mobile.
Obviously it didn't work out, which perhaps makes your point even more true.
<pedant>
Alligators and crocodiles are crocodilians, they lived alongside dinosaurs but are a distinct group.
</pedant>
It's hard to predict but even code and concepts from Lyons LEO, ICL.. live on in todays tech world so I would say that provided we don't destroy the planet Windows will live on and evolve rather than die out. Whether its AD, or windows server core influencing linux, bsd or something else or whether the client retains widespread use, that I don't know.
There is plenty of evolutionary pressure in the data centre AI, HPC, and other intense workloads still want more power more compute more GPU.... but on the other end operators want commodity hardware and appliances to use less enrgy and run colder but deliver the same capability as before. On the client side people expect mobile devices to just work on cellular and wifi battery life to be 10+ hours and access to the same files across multiple platforms with as much or as little encryption as they choose. More and more they don't want to throw away working hardware they want mobile devices to connect to their Car, TV, door locks and wherever else. Whilst for now many are blissfully unaware of the security omissions and messes this is bringing when that becomes an issue the expect it to be fixed with an update. Not only Microsoft face headwinds in that climate. Laptops and Desktops are expected to work as before but connect to all of this and run colder with less power. Most end user don't want client side AI requiring them to carry a petapascal network in a very heavy latop bag or at least many terabytes of data and a CPU/GPU that is overkill for every other workload.
Unless AI comes up with some USP everyone will need to evolve when the AI bubble deflates.
"It's hard to predict but even code and concepts from Lyons LEO, ICL.. live on in todays tech world so I would say that provided we don't destroy the planet Windows will live on and evolve rather than die out."
Windows had been built around a single platform. Unix was built with multi-platforming in mind. It's not surprising that OSes with a Unix background (I count Apple's offerings in that) find it easier to adapt.
That's simply not true. Windows has been run on all of the following at some point in history: x86, x64, ARM, PowerPC, and I'm pretty sure DEC Alpha too. I've used all but the last.
MS first showed off Windows running on ARM back in the 2010s (2008 if memory serves); they just did the HAL, built everything, and had Windows 7, Office and an Epson printer working at some trade show.
Windows' long run on Intel has been remarkable, but is in no way due to limitations in its architecture.
The hardest port is the first one.since Win11 is actively supported on x64 and ARM, I would expect no serious problems porting to a third architecture. Quite a lot of the kernel (is the HAL still a recognisable thing, idk?) might need attention. Hardly anything in user space would, even if it is still mostly C++. (Other languages make it quite challenging to create a portability issue even if you wanted too.)
Most of the problematic stuff has already been fixed when they made modern Windows run on ARM, which it does well. By now, they've identified and likely removed most things that might have crept in that's X86-specific. That doesn't mean porting it to RISC-V is just changing a compiler setting, but it's a lot easier than you're claiming. And before we pat someone else on the back, check what adding RISC-V support to Linux looks like. It's not simple there either. There are lots of patches, mailing list fights, regressions, and manufacturer-specific editions to get there. Linux is ported to more things because it's used on more things, not because it's got the architectural trump card that makes it easy.
>I thought someone might come up with that but it's a long time since it was developed for those architecures.. How easy would it be to take W11 now and port it to them?
One way or another, Windows has been on ARM continuously for nearly 2 decades in one form or other (demo, WinPhone [the later proper NT kernel versions], tablet, full desktop machines).
Windows 11 is interesting because it does have vitualisation based security, though this can be turned off. Processes are run inside a VM, not just as a process. VBS requires hardware support, though most major architectures have variations of the basic ideas. VBS is on Windows on ARM, so that's pretty much all of Windows 11's full gamut of features rolled out across ARM, as well as AMD and Intel (I think there's difference between those two when it comes to virtualisation, secure enclaves, etc).
and like the dinosaurs it will have to evolve to survive or die off (eventually).
It has evolved, just in the wrong direction and needs a major slapdown from the regulators.
Back in the good'ol days it was mostly an OS. We could run apps if we wanted to. They'd load, we could do our thing and then close them and do something else.
Win11 is less of an OS and more a data harvesting and monetisation tool. Even when we pay for a licence. We're going to get Copilot and probably Totalitarian Recall, whether we want it or not. Then for power-ish users like me, it's a constant task of going into Task Mangler, finding carp I don't need and trying to remove or disabling it.
One of my favorite examples is this-
Microsoft Phone Link (formerly Your Phone) is a Windows application that synchronizes your Android or iPhone with your Windows PC, allowing you to manage calls, texts, notifications, and photos directly from your computer.
I don't want to do this, I never gave MS permission to do this, but MS really wants to slurp calls, texts, notifications and photos because the damn thing keeps coming back. And it's consuming 46.5MB of memory, and I yearn for the days when a PC could run an OS and apps in only 640KB. But I like this one because I think it's symptomatic of the rot & abuse that's deeply entrenched at MS.
So Task Mangler lets us view Startup Apps. Phone Link is in that list, and status is disabled. That should be a strong indication that I don't want, or consent to it running. And yet it is. And if I end the process in Task Mangler, that's another consent problem because after a few minutes, it'll be back again. Plus there's another process group for Mobile Devices hogging another 13.2MB.
Why?
Especially when the UK Computer Misuse Act makes it illegal to access or modify computers without the owners consent, and that trumps MS's EULA.. Maybe. Luckily tools like this exist-
https://www.oo-software.com/en/shutup10
that allow users to take back some control, but it shouldn't be neccessary.. Especially if you look at the long list of things that can and arguably should be shut up to increase user's security and privacy.. But most users are probably blissfully unaware of the potential problems, and MS likes to keep it that way.
<< So Task Mangler lets us view Startup Apps. Phone Link is in that list, and status is disabled. That should be a strong indication that I don't want, or consent to it running. And yet it is. And if I end the process in Task Mangler, that's another consent problem because after a few minutes, it'll be back again. Plus there's another process group for Mobile Devices hogging another 13.2MB. >>
Yet another reason to dump Windows.
"The Linux fetish for continuity..."
You make it sound like it's nothing all that important. Bovine manure! It is essential in any self-respecting OS.
The purpose of an OS is to make hardware usable. If a new version of the OS renders said hardware unusable. that means the OS is unsuitable for its purpose.
W11's inability to run on older but perfectly good hardware after W10 has abandoned those who paid good money for it shows how Windows does not enable productivity but inhibits it. Putting continuity before all else is not a fetish. It's the only sensible approach to OS maintenance.
I didn't read that line the way you did. I suppose I take a rather old-fashioned dictionary approach to fetish as an obsession. and the reason it is an obsession is because it is common sense. As a result Linux doesn't have problems and doesn't even have to resort to abstraction layers to provide cross-platform portability.
For many of us, our digital environment is primarily Arm or AMD, and a Linux or Unix-derived OS.
I suppose that depends what you consider to be the 'primary digital environment'. If it's a mobile phone or other consumer device then fair enough. But if it's a desktop then '..Linux or Unix-derived OS.' is an overly optimistic view of the market.
Microsoft's fate, presumably to collapse or to suffer disaggregation of viable components, seems to follow a natural pathway for constantly innovation-dependent businesses. What's notable about Microsoft is the pace of its rise and decline. The decline is partially occulted, but underlying structures and processes make it inevitable. The, in historical terms, near frenzy of Microsoft may have resulted from particularities of the present era: rapid technological advances; huge demand for products relating to computation and communication; relevant physical devices fast declining in unit cost; easy global marketing reach; and competition which could not indefinitely be kept at bay. Adding to that, the decline of initial entrepreneurial zeal has resulted in a colossus led by the blinkered vision of money-men and business school graduates: people lacking the emotional equivalent of 'skin in the game'.
BOEING, founded in 1916, illustrates a company suffering similar straits, but arising over a protracted period; the merger in 1997 with McDonnell Douglas being a failed attempt at revivification. BOEING, at one time driven by engineering talent, and Microsoft, initially dependent upon software developer input at the highest level, have succumbed to routine adherence to mantras issuing from business schools; these schools themselves besotted by the market-capitalism destroying doctrines of neoliberalism.
By contrast, makers of always in demand physical products deemed 'widgets', e.g. nuts and bolts, may introduce little mindboggling innovation during the course of two centuries.
Banking, a not intellectually taxing business when run with honest intent, has barely altered during many centuries. However, its recent manifestations have adopted practices and technologies anathema to prudence and decency; yet, these were not entrepreneurial because bank CEOs are nothing more than employees of shareholders able to walk away with rich pickings regardless of their deceit, incompetence, and collapse of the bank employing them.
Perhaps most business can be categorised according to the three kinds exemplified here: the initially innovative but inevitably running out of steam; the staid, reliable, the producer of routinely in demand goods and services; and financial infrastructure business enabling its grandees to place themselves in undeserved positions on pedestals. Running alongside banking are 'intellectual property' rentier businesses: the similarity resting upon each indulging in selling monetarily worthless pseudo-products e.g. Hollywood films, and rehypothecated gold.
In the last century, there was AT&T (now a modem test command). It got its vibe going allowing "universal service" back in the 1910's (as I remember it). This was the beginning of its downfall which happened in 1984 with the breakup mess. Later the result was two companies (AT&T and Verizon) that now do most of the telecom work in the USA.
Windows (Microsoft) and Intel had a symbiotic relationship that started with MS-DOS, and has continued to Windows 11. But that is starting to break apart. People are seeing alternatives to this structure much as long distance carriers did with AT&T (note: Long distance subsidized local service by a bunch). Now people are adapting and the crack are starting to appear. Intel is floundering (government getting a chunk of it), and Windows 11 doesn't seem to be "better" (for some value of "better") than its predecessor.
Life goes on. Those who don't study history are doomed to repeat it.
Well.......I'm running Fedora 42 and WINE 10.
....and this setup is good for some pretty old application licences:
- M$ Multimedia Viewer (Windows 3.1, 1994)
- MS Visual FoxPro 6 (Windows 98, 1998)
- Quicken 2005
Now.....I don't (and won't) use Windows 10 or Windows 11..........but I'm prepared to bet that Fedora/WINE provides MUCH more backward compatibility than anything Redmond can offer!!!
Strange too that commentards here simply don't mention WINE....I wonder why!
Similar (in my experience) with some legacy Windows games.
I've had issues in Windows 10 with games like KotoR (1 & 2), Sword of the Stars (1st version), CnC Generals etc. They either don't run at all in Win 10, require tweaks to get them to run (like hunting down ancient runtimes or dll's), or they run but then have glitches, or worse are a bit crash happy.
But under Linux, via Proton (which of course is a Valve (and other contributors) modified version of WINE focused on gaming), and they work fine.
Apple gets a pass for generously keeping 2017 era Macs alive till 2027 because they had a platform transition, but Windows 10 from 2015 terminating in 2025 is unacceptable? And windows supports multiple (some) platforms anyway?
"Macs from 2017 will finally lose support in 2027 – although if you were buying an Intel Mac in 2021, it'll suffer the same fate"
Microsoft's 10 year support lifecycle has been a thing for a long time (just don't mention Sharepoint 2019), so it's not like it's a surprise. I remember being left high and dry by Apple on an original iMac after not very long. If people were still shipping non-TPM hardware in the recent past shouldn't they be getting a kicking?
This post has been deleted by its author
If people were still shipping non-TPM hardware in the recent past shouldn't they be getting a kicking?
I think that's a valid point. I'm typing this on an HP Probook 470GS that I bought in September 2018. I doubt it was leading edge then as I was only replacing my 'daily browser'. Yet here I sit running Win11. It upgraded to it automatically shortly after it went GA.
So my at least seven year old laptop is running a fully supported version of Win11. Why are there machines younger than mine (and lets face it 7 years is old for a computer) that can't be upgraded to Win11? Who's fault is that really?
Now where I could start to be irritated is if my laptop can't run Win12 due to Co-pilot. TPM is a reasonable security feature so I think it fair to make it part of the minimum spec but AI is just a load of arse.
>but when the progress is so insubstantial
What progress?
Win11 is just Win10 with more-aggressive antihuman work.
Same interface but with even LESS ability to do anything or even SEE what's happening/doable. Plus even MORE pure garbage. And I haven't yet seen a single additional feature. (Other than some microscopic underthehood trivium that in the old days would have struggled to rate even a mention in the release notes)
I don't disagree on the lack of useful new features in 11. I suspect they might have thought about calling it 10x or something, but then they'd have had to support 10 indefinitely, so the only way to kill off non-TPM systems and get the security uplift was to do a major release and wait for the 10 years to expire. How much more secure things are when you still have users is another question entirely.
But it's not like they charged for the upgrade, it was trivial to get it for free. Now making it "Windows Subscription Edition" and charging $80 a year or something would have probably caused a major upset. Wouldn't bet against it though. Virtual desktops take you that way.
Whilst this is possibly true the end-user OS is lost in the noise.
Cloud computing is king and with the market dominated by a very small number of US based corporations it is almost too little too late to start talking about data sovereignty. That critical piece went out the window when the first customer's outside the US started signing up and drinking the Cloud Kool-Aid.
From a consumer perspective there really is little hope. There is an insignificant minority that don't have at least one of the following:
Google Account (Android)
Apple Account (iCloud & iOS).
Meta Microsoft
Amazon.
I would wager a very large cake that a small number of the users care about data privacy and and almost none data sovereignty.
Consumers want convenience and care little about anything else. If you can swap your device without having to actually copy data then it is a win. This is fundamentally what all the consumer use cases are based around. That this is also becoming normalised for corporate use is another matter. That is fundamentally driven by CFO and the board because they want to move the capacity costs of DCs into revenue costs for an M365 subscription or a cloud fileshare.
For corporate customers it is a completely different think however many are so far down the cloud route then cannot back track. Just replacing M365 is a challenge as what do you replace it with?
The only viable alternative is Google Workspace and that is the same issue.
The easy bit is replacing the OS & basic Office functions, the rest is a complete ball-ache with huge costs in both developer and hardware. If you replace Teams with Slack have you gained anything? It is owned by SalesForce, yet another US computing monolith.
It is also completely pointless to roll you own if it is still in someone else's cloud. That means you need data centre space with all the associated costs to get running.
Now as I'm an IT geek, probably not a typical sample, but my household looks like this:
- 2 Android phones
- Android Tablet
- Chromebook
- GNU/Linix Desktop (Mint)
- The router is either Linux or NetBSD IINM
No Windows, and only the Chromebook is Intel (and it could have been ARM). At $WORK it's an M2 Macbook. My S.O. has at their work a Windows laptop, but their work-phone is an iPhone.
I've argued in another article that Intel hasn't Innovated since the 1970s. As for MS, well, we all here know the history of that.
It's businesses keeping Wintel alive, more through inertia than choice.
I'm hoping the future is something UNIXy on RISC-V ;-).
"The stench from this decaying pile of dogshit is finally driving even some corporations away. About time.”
Although I can’t disagree with your comment, I do fear it has a degree of wishful thinking. Yes there will be stories of some companies looking elsewhere, and they may be looking towards Cupertino rather than Linux, but over all, alas no, the vast, vast majority of corporates won’t be dumping Windows anytime soon - they can’t.