News for hackers. Stuff that matters.
Re:MinC Is Not Cygwin by thingfish on Wednesday April 23, 2025 @06:52AM (#43770617)
You're right. I'll change it to "Windows XP or later".
Re:MinC Is Not Cygwin by bandrami on Wednesday April 23, 2025 @07:36AM (#43770909)
IDK, I wouldn't be surprised if it ran on Windows 2000 (still the best OS Microsoft ever made)
Re:MinC Is Not Cygwin by guerrilla on Wednesday April 23, 2025 @08:48AM (#43771505)
What's wrong with 2003?
Re:MinC Is Not Cygwin by bayindirh on Wednesday April 23, 2025 @10:00AM (#43772318)
Windows 2000 was not developed for x86 directly. It first started on Alpha, then ported to x86, so its codebase is "double-brewed" in a way.

I also remember 2000 as the only rock solid Windows release, and I never had to reinstall it. XP was very close to that after it fully matured, but nothing feels like 2000.

[0]: https://en.wikipedia.org/wiki/Windows_2000

Re:MinC Is Not Cygwin by adra on Wednesday April 23, 2025 @10:25AM (#43772611)
The two OS kernels and API are super close (outsider perspective). I used win2k for like 10 years mostly on the back of applications and games supporting XP for so long. I can't recall the big API differences. Maybe XP has UAC and there were APIs to check for it? Anyways, I still have fond memory of manually patching out API calls hard coded into EXEs to bypass XP only parts which were almost always superfluous.
Re:MinC Is Not Cygwin by bayindirh on Wednesday April 23, 2025 @10:39AM (#43772761)
Yes, I'm aware. Normally they shouldn't behave wildly differently, but the new GUI and userspace features made the system a bit flaky at first. Also, XP became a bit more finicky about drivers over the years, and this broke stuff, too.

2000s kernel was one of its better features, but lack of blingy GUI was also helping in its stability IMHO.

Re:MinC Is Not Cygwin by chungy on Wednesday April 23, 2025 @04:31PM (#43776341)
That's not really quite accurate. Windows NT began life on the i860, and then was ported to i386 and MIPS, and very shortly after the initial release, ported to Alpha. A year after that, PowerPC.

Through the life of Windows NT 4.0's life span, MIPS and PowerPC died early, and Alpha support was axed just before the very end. In Windows 2000's development, i386 and Alpha were the only ports left, and Alpha was axed before it could make it to final release.

x86 and Alpha lived simultaneously for most of its development. It wasn't done "first" on Alpha (quite the opposite).

Re:MinC Is Not Cygwin by creatonez on Thursday April 24, 2025 @04:19AM (#43780380)
True, but the flexibility that two attempts at porting to botched architecture introduced also resulted in 2003 becoming Microsoft's first x86-64 operating system. The very next day after the 2003 release, they released Windows XP Professional x64 Edition based on the 2003 codebase, which a user could run on their brand new x64 computer without much of an issue -- that is, if they had even heard of it... it was under-marketed because they needed to save the hype for Vista. So the struggle and failure may have been worth it in the grand scheme of things.
Re:MinC Is Not Cygwin by tomrod on Wednesday April 23, 2025 @09:11AM (#43771752)
It's worse than 2000, IMO, bloated and not super stable like 2000.

Using Windows 2000 is like using a well-crafted Linux distro. Things just work.

Re:MinC Is Not Cygwin by yellowapple on Wednesday April 23, 2025 @10:10AM (#43772455)
If Microsoft had just iterated on Windows 2000 forever instead of continuing with XP/2003 and beyond it's quite possible I wouldn't be a Linux desktop user today.
Re:MinC Is Not Cygwin by actionfromafar on Wednesday April 23, 2025 @09:09AM (#43771729)
A bit bloated. I think NT 3.51 is the coolest, extremely stable.
Re:MinC Is Not Cygwin by chuckadams on Wednesday April 23, 2025 @09:41AM (#43772078)
The NT 3.x series might have had stricter userspace separation than 4.x did where they moved things like GDI into kernel space, but it still fell over hard at even the faintest whiff of fuzz-testing the native API. It wasn't until 5.x (Win2K) that they started taking that sort of thing seriously.
Re:MinC Is Not Cygwin by heavensteeth on Wednesday April 23, 2025 @08:48AM (#43771509)
right, I suspect "NT" is the category they're looking for.
Re:MinC Is Not Cygwin by tomrod on Wednesday April 23, 2025 @09:10AM (#43771748)
Hilarious that this was literally my exact thought.
Re:MinC Is Not Cygwin by haglin on Wednesday April 23, 2025 @10:28AM (#43772642)
Yeah, Windows 2000 was probably the best operating system Microsoft ever produced. It had no crashes, a snappy/intuitive/sleek user interface, and DirectX support for games. I am currently running Mac OS X because I never liked Windows XP/Vista/10/11.

In hindsight, the progress Microsoft made in the ten years between Windows 3.0 (1990) and Windows 2000 seems incredible. They transitioned from a 16-bit shell running on top of DOS to a fully-fledged operating system with a hardware abstraction layer, preemptive multitasking, permissions, a registry, proper device drivers etc.

Twenty-five years later, there has been basically no progress.

At a minimum, I would have expected something like the Knowledge Navigator

https://www.youtube.com/watch?v=umJsITGzXd0

Re:MinC Is Not Cygwin by chungy on Wednesday April 23, 2025 @11:07AM (#43773075)
> It had no crashes

Windows 2000 is my favorite release ever, but anecdotally, I can deny the "no crashes" claim. Sure it was rare, but I had my share of BSODs and weird behavior on it. Still much more solid than... pretty much every other release.

Re:MinC Is Not Cygwin by jug on Wednesday April 23, 2025 @12:42PM (#43773987)
I think since Windows 7, it has been very solid except for maybe third party kernel driver issues which needs direct kernel access out of necessity. Today when you have a BSOD, it’s like a Linux, Mach kernel panic etc - 99% likely to be a hardware problem.
Re:MinC Is Not Cygwin by floxy on Wednesday April 23, 2025 @02:30PM (#43775205)
CrowdStrike
Re:MinC Is Not Cygwin by SkiFire13 on Wednesday April 23, 2025 @02:39PM (#43775303)
> except for maybe third party kernel driver issues which needs direct kernel access out of necessity.
Re:MinC Is Not Cygwin by fsckboy on Wednesday April 23, 2025 @11:59AM (#43773606)
>In hindsight, the progress Microsoft made in the ten years between Windows 3.0 (1990) and Windows 2000 seems incredible.

compared to what, Linux? the BSDs? Solaris? OS/2 didn't achieve market success but it added major feature subsystems at the same brisk pace.

Re:MinC Is Not Cygwin by Borg3 on Wednesday April 23, 2025 @11:38AM (#43773379)
Yeah. It was indeed great, I was using for 10 years, then moved to Win2003 with I still use. But I think Win2003 is even better. It have better kernel, proper SMP support (multicore), good PAE support (16GB RAM), newer drivers, etc :)
Re:MinC Is Not Cygwin by Aloha on Wednesday April 23, 2025 @12:05PM (#43773660)
Why use something so… old?
Re:MinC Is Not Cygwin by giancarlostoro on Wednesday April 23, 2025 @12:20PM (#43773824)
They want to see if anyone will ever 0-day their system.

To be fair, I have old MacOS laptops that are my wife's laptops, we keep them around because they still turn on, and they have all her old files (I've backed it all up several different ways for her), but I never go on the web with any of the browsers installed on those systems.

Re:MinC Is Not Cygwin by chasil on Wednesday April 23, 2025 @02:01PM (#43774870)
Fans of SimH might give you many reasons for enthusiasm in retro-computing.

https://simh.trailing-edge.com/

Re:MinC Is Not Cygwin by Borg3 on Wednesday April 23, 2025 @04:15PM (#43776201)
Because its snappy, have great GUI, I control it, I have all the tools for maintenance (compilers, DDK, disassemblers)..

I know, there pretty nice Linux distros that are lightweight and snappy too. One day... :) but not yet :)

Re:MinC Is Not Cygwin by musicale on Wednesday April 23, 2025 @10:59AM (#43772983)
> a fully-fledged operating system with a hardware abstraction layer, preemptive multitasking, permissions, a registry, proper device drivers etc.

Windows NT, which was developed by people (Dave Cutler et al.) who knew how to design and build operating systems (VMS, etc.).

Re:MinC Is Not Cygwin by fuzztester on Wednesday April 23, 2025 @05:12PM (#43776665)
there is a book about it, called "inside windows nt". I had read it back in the day,

it is by helen custer. it talks about dave cutler's work on it, apart from other stuff.

Re:MinC Is Not Cygwin by p_ing on Thursday April 24, 2025 @01:26PM (#43785218)
Got a copy sitting right next to me! Along with Inside The Windows NT File System (and The Design and Implementation of the 4.4 BSD Operating System).

'Windows Internals' series is the successor to Inside Windows NT.

Re:MinC Is Not Cygwin by fuzztester on Friday May 02, 2025 @08:10PM (#43875744)
ha ha, mark russinovich is one smart dude. and so is charles petzold before him.
Re:MinC Is Not Cygwin by cgannett on Wednesday April 23, 2025 @01:24PM (#43774457)
I had a weird old "laptop" that I was given as a kid, didnt know jack about computers back then so I couldnt tell you what it was. It was one of those that came with a big heavy docking station and it couldnt run without it anymore so I assume the battery was dead. Had Windows 2000 and I used to play AoE 2 on it all day and it was smooth and the interface was snappy. Fond memories. I remember being really confused when people told me that Windows 2000 was bad.
Re:MinC Is Not Cygwin by chasil on Wednesday April 23, 2025 @01:55PM (#43774800)
Bear in mind that Microsoft was the largest commercial UNIX vendor in licensing XENIX for the TRS-80 Model 2 alone. From this experience, they had significant knowledge of 32-bit preemptive multitasking.

That exposure allowed them to see how valuable Dave Cutler's PRISM team would be in a total redesign in melding VMS and (whatever tolerable) UNIX features into a new kernel, but focusing it upon Win32.

There were OS/2 and POSIX emulation layers also, but these were obviously second class.

Re:MinC Is Not Cygwin by signal11 on Wednesday April 23, 2025 @03:09PM (#43775631)
I love Windows 2000, it was pretty great, but display driver crashes would take down the system (similar to NT4). Also if 2000 was installed in most homes it’d be about as much of a fiasco as pre SP2 Windows XP. IIRC there were a bunch of worms that did affect Windows 2000.

Windows 10 on the other hand will deal with display driver crashes just fine, falling back to safe mode.

Re:MinC Is Not Cygwin by zoobab on Wednesday April 23, 2025 @08:10AM (#43771173)
"later" is for lazzy developers.

Better list the complete list of versions where it has been tested, like "Windows XP/7/8/10/11"...

Re:MinC Is Not Cygwin by zoobab on Thursday April 24, 2025 @03:32AM (#43780127)
The frontpage says now:

"It runs on Windows NT, version 2003/XP/Vista/7/8/10/11."

Victory!

Re:MinC Is Not Cygwin by tomrod on Wednesday April 23, 2025 @09:10AM (#43771742)
Win2k?
Re:MinC Is Not Cygwin by dimava on Wednesday April 23, 2025 @11:03AM (#43773020)
Also you are confusing Visual Studio and VSCode