/sci/ - Soyence and Technology

Amiga won everything it could won

NT dontesn't run on non-x86 though

>>25169
NT runs on:
- i860 (at some point in development)
- IA-32 (PC, UEFI, PC-98, SGI, etc.)
- Ijakium
- x86_64
- MIPS
- PowerPC
- Alpha
- ARM
- ARM64

>>25173
runs ran

>>25173
One version for all and all are defunct except 11 for ARM64 and 11 for ARM64 runs like shit on everything but a VM for an Arm Mac (A Mach BSD UNIX, btw)
>durr ia-32 and amd64 are different platforms
No, they're not. AMD64 is binary-compatible with IA-32.
I only count 'run' as being usable. That disqualifies ARM32 from the get-go (WinRT)
All the others are some obscure-shit NT4 build that didn't even get service packs, or in the case of the itanic, an unupdated 5.1 and an unupdated 5.2 build. Both are not usable, and if you got a modern PowerPC system for example, you, de facto, would not be able to run WIndows on it. This goes for Arm64 systems, even, on everything but MS hardware and VMs. How's bootcamp for Apple Silicon going?
So no, NT is not cross-platform in practice, it is x86-centric.
Also, OS/2 and DOS are MS products too while NT was developed by transplants from DEC's VMS, a minicomputer vendor that was also responsible for UNIX, and the early beginnings of Linux through the Alpha (that Windows NT dropped after 4.0 btw), the C:\ stuff is designed to look like DOS (TENEX-like)
Mac, isn't listed, neither the classic Mac OS or NeXTSTEP/MacOSX either, so NSGODS stay NSWINNING

>>25175
Nobody uses any arch other than x86 and sometimes ARM64 for the desktop. It literally does not matter. CE was the offering for embedded platforms, not NT.

NT 5.x and earlier requires an ARC compliant firmware. Yes, even x86 NT required ARC compliance; NTLDR and NTDETECT were shims to emulate ARC firmware.

NT 10 will run on most x86 or ARM systems assuming out of the box, assuming that they have UEFI and required drivers. Linux on the other hand relies on the retarded static device tree system which means that you have to manually define the hardware, not plug and play at all. With UEFI, ACPI takes care of this automatically. This is also why Windows 11 was able to be ran on over two dozen cell phone models.

>>25173
Certainly not as lightweigth and smooth as a UNIX system tho

>>25186
IRIX and NT have similar performance I believe. Your operating system has very little to do with how your software runs. It’s mostly dependent on your compiler, libraries, and hardware.

NT WON
CHADNUX WON
EVERYONE ELSE LOST.
I have a laptop that runs windows and a PC that runs Ubuntu

>>25208
tsmt

>>25177
>Nobody uses any arch other than x86 and sometimes ARM64 for the desktop. It literally does not matter. CE was the offering for embedded platforms, not NT.
fuck you PowerPC is a good platform
>NT 5.x and earlier requires an ARC compliant firmware. Yes, even x86 NT required ARC compliance; NTLDR and NTDETECT were shims to emulate ARC firmware.
holy niggerald, this is why people don't like NTOS, it has hacky shit like this
>Linux on the other hand relies on the retarded static device tree system which means that you have to manually define the hardware, not plug and play at all.
which means your shit won't break because of ??? ACPI device manufacturer bullshit and bluescreen which is what NT will do everyday

Also CE is dead, cause MSFT wants you to use Linux on those platforms, right? Why would I give a shit about an OS made by people who could not give less of a shit if you switched away from it, and actively do so themselves? NT is a fucking shitbox these days lmao

>>25366
>fuck you PowerPC is a good platform
Why?
>holy niggerald, this is why people don't like NTOS, it has hacky shit like this
Hate to break it to you but Linux on non-PC platforms uses hardcoded device trees.
>which means your shit won't break because of ??? ACPI device manufacturer bullshit and bluescreen which is what NT will do everyday
A device tree build of Troonix is only compatible for that EXACT configuration of hardware. Change the motherboard? Your OS won't boot without a full recompile.

>>25389
>A device tree build of Troonix is only compatible for that EXACT configuration of hardware. Change the motherboard? Your OS won't boot without a full recompile.
If it boots it always boots, which is better than Windows, where it has to rebuild its device tree on install (retarded microkernel things), even activation is tied to how it builds its tree (HWID, it's why you have to reactivate Windows if you change mobos). If you swap a general Linux across motherboards it'll not skip a beat cause it already has its trees for all your boards compiled while NTOS freaks the fuck out and basically has to re-windeploy itself like you installed from an environment, which is riskier on an existing install, because NT DOES use static trees, it just builds at install-time, not compilation, which is more fragile. For gods sake a CBS Windows update rendered my laptop unbootable, hung on boot because iRST Intel RAID controller drivers are baked into the NT table permanently on that install because NT fucking glues itself to your hardware on deployment in a way Linux doesn't (because you can just recompile right, no need to tamper with crap on runtime like Windows loves to do). It's so bad that on systems with Modern Standby (that bullshit alone means NT lost, why is the kernel trying to put the CPU in a sleep state? That's what S3 is for.) if you swap it to a conventional power model Windows won't even get to deploy and it'll just panic and bugcheck (Same thing as a kernel panic from a Linux device tree issue, but when you compile the kernel and it boots you have a guarantee that it won't do that in a way NT doesn't offer)
It's even worse on Arm64, if you resize the root partition on NT for Arm64 you also end up rendering it unbootable with blue-screen bugchecks. I know because I toasted multiple VMs that way
>Why?
Proper firmware, and high-end servers use it. Open standard in a way RISC-V and its extension bullshit isn't, too.

>>25414
>Proper firmware, and high-end servers use it. Open standard in a way RISC-V and its extension bullshit isn't, too.
You bitch about ARC but will sing the praises of OpenFirmware? I sure hate things working! Make sure you get the exact correct model of video card and exact right model of IDE card or your computer won't boot. And you said PowerPC, not Power ISA. PowerPC is the pre-2006 incarnation of POWER. PowerPC has horrible compatibility between CPU releases. You can't run BeOS on a 750 but you can on a 604. High End Servers also run x86 and ARM. It's just IBM's midrange systems that run Power.

>>25418
I bitch about NT requiring ARC firmware on MBR systems, schlomo.
> PowerPC has horrible compatibility between CPU releases.
So does ARM, Arm (((SoC)))s boot different from each other, even. I hate SoCs btw, not even real computers just use a normal CPU
x86 firmware is a godawful joke these days, EFI is more bloated than even Apple iBoot garbage and all x86es are ABSOLUTELY PROPRIETARY, even with LibreBoot cause of the ucode and embedded controller shit
MIPS, SPARC, Alpha exist too. All architectures that are perfectly good and not supported by wintard NT because they're intelfags

>>25518
All enterprise Power systems have OOBM. SPARC also does. SPARC isn’t supported by anything. You can “run” Debian on SPARC64 but it’s all statically linked for no reason, missing libc, stable release is broken but unstable isn’t, and APT is mostly broken. All CPUs are proprietary will the sole exception of the OpenSPARC T2.

>>25518
>I love register windows
Are you brain damaged?

>>25526
Btw, by the only CPU I mean only physical 64-bit CPU.

>>25526
>All enterprise Power systems have OOBM
Not niggerfirmware like (((UEFI))), especially no sleepfag mode, and DEFINITELY not (((Modern Standby)))
You never explained why (((always on, always connected modern standby))) isn't an immediate fail for Windows, either. Fuck standby, just turn your PC off.

>>25531
Because your computer starting up faster is good, actually. SPARC T2 hardware runs an ancient version of Linux on it's iLOM with an ancient version of OpenSSH full of vulns and uses an encryption alg so old you have to re-enable it with a flag in ssh. The iLOM, unlike UEFI, is always running and is on a separate ARM CPU. Same goes with OpenBMC on most enterprise servers. You probably don't even own any SPARC, SGI, or Power equipment. Clearly proven by your attacks against ARC even though ARC is a MIPS standard, the big endian version, ARCS, being used in most SGI computers.

>>25535
>Clearly proven by your attacks against ARC
RETARDERALD. I have NO qualms with ARC, I have qualms with NT using IRIX's boot-structure on fucking MBR DOS machines because that's a sign of bad quality code. I have 0 issue with an IRIX box booting like an IRIX box. I have an issue with Windows NTarderald booting like an IRIX box on a system that hasn't ever seen a Unix-like in its life and would be booting to a C:\ prompt if it weren't for a bunch of retards trying to shove an overglorified minicomputer OS on a processor that can't catch up with itself enough to blink a text cursor.
>Because your computer starting up faster is good, actually.
I'd rather wait 5 more seconds for startup than have Windows PE headers in my firmware and a giant code hairball that WILL get me hacked
>SPARC T2 hardware runs an ancient version of Linux on it's iLOM with an ancient version of OpenSSH full of vulns and uses an encryption alg so old you have to re-enable it with a flag in ssh.
EFI runs a version of MINIX on Intel. FUCKING MINIX, FROM THE 80s. MINIX! CAN YOU EVEN FUCKING INSTALL MINIX ON YOUR MAIN CPU? Because of course we're bringing out-of-band management solutions in and conflating them with the boot-process, I can do it too with the Management Engine.

>>25535
And for the love of god, (((always on always connected))) doesn't just make your system boot faster, it also means your fucking shit doesn't have a real ACPI sleep state, the CPU is forced on by the OS, it can wake over anything and waste power, and niggerupdate on you while in a bag. Fuck modern standby.

>>25705
Using a single system with veneers simplifies the boot process.

>>25721
>Using a single system with veneers simplifies the boot process.
No boot process is as simple as MBR.

all of you are retarded, NT is a decent OS, so is Berkeley UNIX, Linux, TempleOS, Solaris, NeXTSTEP and whatever the hell else you use as long as it's not DOS. Could be a damn LISP machine or something and it'd still be good cause it's an actual multitasking proper OS.

>>25724
WTF are you talking about. "MBR" isn't a boot process. It's a partitioning system. ARC NT requires MBR.

>>25733
no it isn't, kill yourself

>>25739
So true, he should have excluded Troonix from the list as it is literal malware.

>>25161
all old news, osakaOS is the next big thing
https://github.com/pac-ac/osakaOS

>>25748
Die tranny

>>25740
>LINUX IS MAL-AAACK
Retard, Microsoft literally uses the Linux kernel on their servers. They also include a GNU userspace server for NT (fucking remember when those were used?) in WSL, if Linux is malware so is Windows.

>>25758
No, Microsoft uses NT for most of its servers.
WSL2 contains the Linux malware in a VM so it cannot damage the rest of the OS with poor design decisions.

>>25763
>No, Microsoft uses NT for most of its servers.
https://www.theregister.com/2023/05/26/microsoft_azure_linux_container/
I sure do love it when 'malware that isn't used on their servers' is used enough that they offer their own Linux distro for Azure. Up there with the time they offered Edge for 'MalwareOS'
>WSL2 contains the Linux malware in a VM so it cannot damage the rest of the OS with poor design decisions.
retarderald. Also WSL1 used a call-translation container (think csrss.exe) for things. And they had a POSIX layer in the 90s too. And they use fucking Linux for Azure. And they contribute code to the 'malware' kernel. Next you're gonna tell me Apple runs the BSD stuff in their Mach systems in a VM because it's "malware" that they regularly commit-dump to. I sure do love when companies commit code to malware they're trying to combat.
Stop drinking the fucking flouride water it's not good for your brain.

>>25766
>Microsoft offers Linux as a service
And all those Linux VMs run on Hyper-V which is exclusive to Windows Server. Yes, I admit that it is very sad that that Pajeet Satya Nutella is pushing malware instead of Microsoft’s bonware (Windows). Ballmer would have put a stop to this nonsense.

Most if not all of the Azure Cloud infrastructure runs on a heavily modified version of Windows NT.
>WSL
I was talking about WSL2 but whatever. WSL1 is not Linux. It’s a ABI and API compatibility layer to run Linux software on NT.

I call Linux malware to make fun of how GNUbrains call software they don’t like “malware” and accuse Windows of doing the same thing that GNU (a division of IBM Corporation) and Linux do. E.g. refusing to follow standards, vendor lock in, corporate control, EEE, anticompetitive behavior, etc.

>>25771
>I call Linux malware to make fun of how GNUbrains call software they don’t like “malware” and accuse Windows of doing the same thing that GNU (a division of IBM Corporation) and Linux do. E.g. refusing to follow standards, vendor lock in, corporate control, EEE, anticompetitive behavior, etc.
imagine being obsessed with cat -v so hard, you try justify the proprietor of internet explorer and WORD

>>25772
>you VILL only have one browser engine choice and that VILL be chrome.
IE was the only thing that kept the Internet from falling into Google's hands. Many do not want to admit this.
>Word
Word is good actually. Sounds like a skill issue.

>>25778
>IE was the only thing that kept the Internet from falling into Google's hands. Many do not want to admit this.
Firefox and WebKit albeit

>>25789
<SAMEFAG
Also I.E. was total garbage until they recoded it to Edge and by then it was too late and Google monopolized the internet so…Yeah.

>>25789
WebKit is the base for Blink and only major use is in Safari which is Apple exclusive. Mozilla is funded by Google.