>too unstable for any production use
>only barely runs on pre-2010 hardware
It's an impressive project but it's still not viable for anything.
I think they started implementing NT 6.0 APIs (Vista) but it's only a start.
Since it's an open-source replication of Windows Server 2003, you can only run it on old hardware from the 2000s, there are no drivers for modern hardware. Otherwise, you must run it from a VM within Windows/Linux, so...

reactos is the poster child for biting off more than you can chew. also they and wine avoid each other like the plague

>>108784539 (OP)
reactos is the buggiest windows alternative, don't run it on modern computers

>>108784567
>>108784564
>>108784561
>don't develop it
>why is it so buggy waaaaah

>>108784594
lil nigga they are trying to reverse engineer windows. doing this legally is an astronomical pain in the ass

>>108784603
nigga we have the decompiled windows xp (aka the best version of windows)

>>108784603
I'd much rather have something that works and not be worried what doesn't work all the time.

>>108784607
You serious?

>>108784607
>nigger doesn't understand law
>>108784620
Yes, it leaked several years ago.
https://linuxreviews.org/42.9_GB_Of_Microsoft_Source_Code_Leaked:_Historicans_Can_Now_Study_The_Source_Code_For_MS-Dos_3.3_To_Windows_XP

>>108784607
It is a cucked project deathly afraid of being sued by MS. They purge devs and code suspected to have influence from leaked code.

>>108784632
Oh, back when Covid was in it's hay-day. I expected that to be released at least a decade before.

Kinda wish there was community efforts into making security patches for Microsoft Defender, but that would prolly be as trustworthy as the W11 Dev Team vibecoding their broken OS

>>108784539 (OP)
"also known as" doesn't mean it's true
hope this helps

>>108784539 (OP)
>can run windows binaries
Can it, though?

>>108784567
>buggiest windows alternative
As opposed to?

>in alpha since 1996
2026 year of the reactos desktop?

>>108784607
If you're talking about the source code leak back in 2020. A. We can't use it legally and B. The source code is still missing source for components anons had to scrape from old versions of windows

>>108784539 (OP)
>can run windows binaries
you haven't even tried it, have you?

>>108784826
Just look at the code and make your own based on it, use AI, all done

>>108784826
Take the leaked Windows XP source code and vibecode all of them using AI, Then import bunch of codes from Proton, Wine forks, ReactOS, etc etc etc into the leaked Windows XP source code repository

>take leaked windows
>claude, rewrite it in rust
>publish
>don't get sued

File: OpenNT_4.5_Demo.png (27 KB, 1024x768)

27 KB PNG

>>108784826
>>108785020
OpenNT is more than enough. Just start from there.

>>108784539 (OP)
I wish it was good, I do not like Unix on a desktop.
What this shit needs is more peripheral drivers and the mesa driver.
Beta quality and And running a lot of software and DWM.

>>108784564
The WINE devs are the ones avoiding any ReactOS code, but they have never explaining their reasoning for doing that.

>>108785796
Linux users are cultists, they look down upon anything Windows-related.
ReactOS devs said the reason they don't "just settle for Linux" is because Linux is fundamentally shit and always will be for home users, see interview:
https://www.youtube.com/watch?v=VnLQvqoxXjA [Embed]

>>108785796
Plausible Deniability you retard.

>>108784840
>>108784840
I ain't got the money for an army of lawyers, nor has any freetard with dreams of making a true open source Windows. Though it will be funny to get sued and lose to destroy every single coding AI because the code it shits out is liable for lawsuits.
>>108785088
OpenNT is just leaked Windows NT 4, no different from the XP leak.

>>108785828
No that's wrong. Like I said they are a cult. They think if you don't use a terminal and bash scripts for everything, you shouldn't be allowed to use a computer.

>>108785844
Not a bad choice. Case in point, you got filtered.

>>108785796
they don't wanna get sued by microshart

>>108785844
Could be. I'm there. I tried Git for Windows at work and now I use Ubuntu at home. Automating things is addictive.

My dad asks for my help using programs. I don't mind, but I also ask myself, "Wouldn't this be easier to understand as a CLI program with a man page?"

>>108785826
It also happens that a lot of devs are unixfans and despise everything Non-UNIX on top of despising Microsoft.
They just don't know how Non-UNIX systems work, or simply don't care.

>>108784603
>doing this legally
did they forget theyre russians?

I fucking hate copyright law

>>108784539 (OP)
Controlled opposition. It has fundamental flaws which are never addressed. It will never be usable because its not intended to be.

>>108784564
reactos could have been done ten times over by now if it was done right. The problem is their entire development philosophy is wrong. The fundamental flaw is developing the entire OS in parallel, this results in a broken eternal alpha state. What they should have done is developed one component at a time, all the way to 100% such that it could be dropped into win2k3 and pass all functionality tests. Also supporting gcc is utterly retarded. Windows should be compiled with msvc.

>>108786770
>What they should have done is developed one component at a time, all the way to 100% such that it could be dropped into win2k3 and pass all functionality tests.
They do that. It's just that they don't have enough manpower to see noticeable progress. And they always test their individual components in Windows Server 2003 first before they merge them into ReactOS.

>Also supporting gcc is utterly retarded. Windows should be compiled with msvc.
They do have MSVC builds, they are on their site. Plus they also have some experimental Clang ones, but I don't think those are public.

>>108784539 (OP)
>>can run windows binaries (aka useful software)
It doesn't, at least not yet aside from very specific stuff.

>>108786716
>It has fundamental flaws which are never addressed

>posts something retarded
>doesn't back it up
Yup, it's /g/ time

>>108786871
>They do have MSVC builds, they are on their site.
He means supporting gcc entails an extra effort that is not worth it.

>>108785355
>What this shit needs is more peripheral drivers and the mesa driver.
They constantly work to improve driver support. Recent builds have vastly improved support for SATA and Nvidia GPUs. Not that many years ago they didn't even have support for USB.

>>108786770
gcc has some features. msvc doesnt have VALs, your statement is repeatedly empty

>>108786871
>some experimental Clang

clang is the top compiler now, everything should be compiled with it, theres some php guy in there.

>>108785826
GNUnix users may be cultists but jeez, those dudes are neurotic laugh reveals them retards, unstable as the OS project itself. found the answer why

>>108786465
Only the founders were Russians, but they have people from around the world as well.

>>108788514
gcc/clang do not support core features that windows requires, specifically SEH. Also gcc/clang are garbage compilers that abuse undefined behavior to make micro-benchmarks slightly faster. I would not write anything important in gcc or clang. MSVC is a professional compiler that is actually trustworthy.

>>108784539 (OP)
My OS choice does not depend on how well it can run closed source dogshit. I rarely need to run windows applications, and if I do wine almost always works.

>>108789832
> specifically SEH
very good, SEH and other telemetry retardation should be dropped. if you want them, install windows 11.

> MSVC is a professional compiler
are you deaf or somth, i said it doesnt support VALs. you cant code anything modern without VALs. nonono, no malloc babaloc substitutes, otherwise install Linux or OpenBSD, it does complete obfuscation and integrated with those libc antiques.

microsoft is moving (moved) to c++ (or rust), msvc is abadonware to thier own norms

>>108789998
Modern Microsoft is not something to look at and emulate. Also, SEH is a fundamental part of windows, it cannot be gotten rid of. Unlike linux which uses buffer copies at the system boundary. Windows passes user buffers directly into kernel mode. SEH is used extensively to make it possible to access user buffers directly without crashing the system.

>>108789832
>MSVC is a professional compiler that is actually trustworthy.
rare anti-freetard dunning kruger larp

>>108790088
MSVC is objectively a better compiler than gcc or clang. MSVC does not hide behind "undefined" behavior as an excuse to break your previously working code. MSVC is not constantly reinventing itself, adding billions of dubious optimization passes that usually make the code worse. MSVC is a reliable and trustworthy compiler, whilst gcc/clang are experimental toys.

>>108790065
> “SEH is used extensively to make it possible to access user buffers directly without crashing the system.” This is misleading. SEH can be used to catch faults (access violations) that occur when kernel-mode code dereferences invalid addresses, but relying on SEH in kernel-mode is neither the normal nor recommended mechanism to safely access user memory. In kernel-mode Windows, there are explicit APIs and conventions (probe/lock pages, try/except when necessary in certain contexts) to safely access user buffers. In user-mode, SEH protects applications from exceptions. In modern Windows kernel development, structured exception handling in kernel-mode is restricted and must be used carefully; using SEH to mask improper validation is poor practice and unsafe.

> SEH is a fundamental part of windows

it is C++ language trait and a runtime feature (exceptions). if you want runtime features, start making a proper runtime first, without those low-level cosplays seen in C++, Rust, Go, etc. so, imo it should be removed.

>>108790132
ok rajesh

>>108790159
> it is C++ language trait
SEH has nothing whatsoever to do with c++.

>>108790132
>MSVC is not constantly reinventing itself, adding billions of dubious optimization passes that usually make the code worse.
That's a roundabout way of admitting it's just less advanced. lole.

>>108790182
If you have ever looked the the assembly generated by gcc and clang you would not be impressed. All the different layers of ultra-complex optimization fight each other to produce some truly retarded assembly. MSVC on the other hand is more conservative, but also never does profoundly stupid things. Its code gen is always ok, whilst gcc clang swing between two extremes, amazingly optimal one moment to profoundly brain-dead the next. Many of gcc and clangs optimizations, turn into anti-optimizations in many cases.

>>108790065
>Also, SEH is a fundamental part of windows, it cannot be gotten rid of.
That's actually a good thing, at least for a consumer OS.
Linux (and other UNIX OSs) expects you and the rest of the devs to do the hard work for them and adapt the apps while Windows keeps all the compatibility configurations since the very first version of the Windows API without changing a thing, or al least keeping changes at minimum, so you could install software from 20 years ago without problems.
With Linux? lol, good luck installing software from 5 years ago, and the only answer linuxfags give you is "compile from source yourself lol".

File: Windows-like_comparison.png (69 KB, 800x437)

69 KB PNG

>>108784564
To be fair Wine doesn't have anything that could be used by ReactOS.
The problem is that both projects are fundamentally different at their core.
ReactOS wants to be an open source 1:1 Windows clone as much as possible, including all the particularities and special cases Windows has.
Wine operates with an "as long as it works" philosophy, accuracy be dammed. Not saying this is a bad thing, but they don't care about being compatible with the OS, just with the software and apps in mind. Besides Wine is heavily tied to Linux, and it's basically its primary OS, so the portability wouldn't be a 1:1 situation anyway.

File: Are you missing how it re(...).jpg (273 KB, 1080x1072)

273 KB JPG

>>108784840
that would immediately cause them to get sued retarded jew. if you want this shit to get gud then donate money
>b-but my return
your return is the product you stupid cheap kike.

>>108790357
>Besides Wine is heavily tied to Linux, and it's basically its primary OS, so the portability wouldn't be a 1:1 situation anyway.
Slowly but steadily they have been adding more abstraction layers so as to not to rely that much on Linux's particularities. For instance, the last version added this:
>More work on MSXML reimplementation without libxml2.
This kind of stuff keeps on happening. I guess they want everything to live in its own little world as much as possible.

>>108784539 (OP)
I don't get what this is. Some kind of windows XP abomination reverse engineered?

>>108784539 (OP)
I just run a Win7 VM for older software I use.

>>108784607
>nigga we have the decompiled windows xp (aka the best version of windows)
it's better to vibe code all of windows xp source code files using ai

>>108790917
Vibe coding with the same code that the AI learned by searching the leaked XP source, lmao

>>108784539 (OP)
ReactOS is a joke. They are trying to copy what Microsoft (a billion dollar corporation) has done over a period of 40 years, with a vastly smaller dev team, and a vastly smaller budget. The only chance in HELL that they can pull it off is if they start using illegal methods such as, at least, throwing their entire project in the trash and using the leaked 2k3 source code as a starting point.
Their insistence on legalfaggotry, despite IIRC all their developers being russians, means they are not a serious project. If you can't break the law, then it's better to just give up, because they're never going to get anything useful out of all this work.

>>108790216
post a benchmark

>>108791340
Most of the code changes Microsoft have made in the past 20+ has been damage and not desirable to replicate. They absolutely could replicate the core parts of win2k3 if they were focused. They have so many advantages that Microsoft did not have. Faster computers for quick rebuilds, decent source control. Virtual machines for testing. The Reactos project has wasted absurd amounts of developer time on bullshit. Like they have a clone of ms-paint for fuck sake. Literally why. They should not have touched a single component outside of the core system, not until core was done.

>>108784539 (OP)
Windows is bad. Everything went wrong with windows 95.

>>108791515
ReactOS should aim for Windows 7 compatibility and that's it. It would be the perfect OS.

>>108790171
__try or try harder to prove it

ReactOS will be the only viable Windows replacement because it will be already fully compatible with windows apps without being in need of tinkering like linux.

Rememebr that:
>Linux is only for enthusiasts because it always require tinkering with it
>All users that daily drive Linux always have a partition with windows installed
>There is no linux distro that works out of the box for something different than browsing the web, use libreoffice and other basic stuff
>Every year is the year of the linux desktop (For them)

The year of the ReactOS dektop is more likely to happen than the year of the Linux desktop

File: r-chamber.jpg (458 KB, 750x1244)

458 KB JPG

>>108790357
>ReactOS wants to be an open source 1:1 Windows clone as much as possible

copycats always fall behind the main thing (if thats true statement). those NODE copycats for example, theres not much sense in re-implementing/re-writing the same retarded JS "standards" with 32-bit integers. theirs is the golem behaviour to get more users, an unexplainable, non-justifiable thirst for users like GNUnixoids have. same pattern with rust re-writers, they are always behind with some unpopular and bogus solution that is line-to-line copy of the thing

the objective must always be of securing superiority over existing products, that creates a certain mindset. just making it work, just compiling, just avoiding the bugs oh no bugs bugs, unsecure, catch them aaa aa.. then you will listen to those neurotic laughter..

>>108784831
No, he read about it, then confidently decided to promote it with bullshit.

>>108791554
Windows 7 compatibility must be carefully implemented, the 2k3 architecture must be retained as much as possible. NT6 was a disaster which we would not want to replicate.

>>108791554
nothing installs on win7 anymore

>>108791795
Not true, and a lot of software is just artificially locked on newer systems. ReactOS just needs to mask itself as a Windows 10/11 system and that's all. That's pretty much what it does now with the 2000/XP/Server versions.

>>108784539 (OP)
vaporware

>not using claudecode to build the bridge between worlds that destroy Microsoft.
>Ngmi

>>108791795
have you ever met a

program.zip

that goes into

program/runit.exe

and just launches? its called portable software. one of the reasons why windows is superior to various NIXes. Zigman tried to change the situation but looks like he failed terribly, because he projected an implicit runtime

How could someone use the leaked Windows XP source code for advancing ReactOS development without legal consequences?
There must be some legal loop that even if Microsoft look at the results and is 99% sure that is copied from them can't win the legal cause.

AI can be viable for this?

>>108791554
Aren't they actually aiming for an NT6+ rebase on the mainline for a future major release, though?
I know one of the core devs was tinkering with this idea on his fork, and them agreeing on eventually moving from NT 5.2 some time ago during chemnitz linux days
>though i guess the nt5 branch would be kept alongside, especially for the xboxhueg and pc98 ports

>>108793816
NT6 was the death of windows. If they go NT6 that is the final nail in the coffin for reactos.

>>108793816
>>108793960
NT 5.2 and 6.0 are irrelevant to them. The ReactOS kernel actually has the functionality of Windows as a whole, the main problem is the functions of NT6+ (Vista and onwards) are harder to implement, so they are focusing on the XP/Server 2003 era.

>>108784539 (OP)
i would main it, trust me i really would
if it wasn't on fucking windows vista standards

>>108794472 (me)
don't mean it in terms of interface (which is on xp standards and id like it to stay that way)
i mean in terms of compatibility, it's generally VERY behind, i'm not that type of unc that exclusively uses mesozoic software. i also mean windows vista standards in hardware compatibility, my laptop is too new for even windows 7 (uefi class 3 damn you)

>>108794472
>>108794490
(both me)
on top of that, it's alphaware, i can withstand beta, sure, not a really raw alpha, i still value stability you know.
either way, i'm satisfied with arch linux + wine on limited and specific occasions

>>108784539 (OP)
>why?
it's not troonix

>>108793208
the legal loop is giving a fuck about legaltards

>>108784539 (OP)
it doesn't work, every time I run it a bunch of windows pop up and then it crashes and says it doesn't work for my system but thanks for trying it

>>108791515
>Most of the code changes Microsoft have made in the past 20+ has been damage and not desirable to replicate
this is outright not true
you need to replicate them so that APPS can run on your OS, you can't just say "well I don't like that part", because then you're designing your own non-compatible OS just like every other tinker tranny with his toy OS.
Sure, the telemetry is not important, the Javascript start menu is not important, but if there's a new Win32 API or even some COM or WinRT nonsense then you at least have to stub it or apps won't even start.
Also around the time of Windows 8.x there were some real improvements in virtual memory handling in the kernel.
>>108794472
It's not on windows vista standards, it's on windows 2000 standards kek

>>108784539 (OP)
It has been in alpha stage for I think over 30 years at this point.

>>108795097
>you can't just say "well I don't like that part"

you can, thats what win7 and other backporters do. in fact its mostly the threading model that causes things to bloat, when you see those popups about app inability to run. obviously, all the frameworks and "studio codes" silently depend on them, upgrade, updoo and this is a race you will never win if you arent steering the whole company's policies.

> the Javascript start menu is not important

its not JS but C++ code, it was pointed out before, but you retards dont like to read those

> Windows 8.x there were some real improvements in virtual memory handling in the kernel.

virtual memory is fully solved with the VirtualAlloc, theres nothing beyong. the swap file is abomination and could be deleted.

https://www.youtube.com/watch?v=c1LjnFKGDhQ [Embed]

https://www.phoronix.com/news/ReactOS-Unified-ISO

>>108789998
Microsoft can't take any meaningful steps to remove old APIs or shift from the C++ codebase. Rust as a language isn't stable, much less has a stable ABI. The day that old Win32 dies is the exact same day that Microsoft dies. They have some cloud infra which could have sustained them. But with the AI bubble popping and cloud costs skyrocketing companies are reviewing on-prem strategies.

>>108790257
That isn't a charitable response to the Linux world. First, Linux never gave the same stability promise. Second, Linux is much more open, so if you have a binary from 5 years ago that you need to run unmodified, telling someone to use 5 year old libs isn't unreasonable - Linux never had the same kind of distribution model that Windows did. Third, Linux is just a kernel and most of the stability issues stem from other systems.

>>108796029
This is actually a good thing. I will try it now in a shitty netbook I have and I hope it would boot it this time.

>>108786716
>Controlled opposition
Opposition to what? And the project is open source, anyone could fork it if they wanted, but the problem is no one wants to do it.

>>108797770
> Opposition to what
A possible legit replacement for windows. This fake project is designed to soak up developer effort whilst being sufficiently mismanaged such that it will never leave alpha state.
> no one wants to do it.
Because the project is such a disaster it might be better to start over than trying to un-fuck it.

File: 1761940820562811.png (21 KB, 865x238)

21 KB PNG

>>108785796
gee I wonder why

>>108784594
Yes. Seethe more.

File: 1752759405786146.png (13 KB, 817x138)

13 KB PNG

another one

>>108797803
>This fake project is designed to soak up developer effort
Doesn't make sense considering this project is just like 20 guys working part-time on it.

>Because the project is such a disaster it might be better to start over
If it's that easy, why no one has done it?

>>108797806
Pure, unadulterated FUD. They even had a third party audit not long ago that verified that no stolen or decompiled windows code was used in this project.

>>108784539 (OP)
what's the point

>>108797827
>In my opinion, the fast call entry code is copied step by step from the disassembled Windows code.
>when I analyzed the fast call code (yes I looked at disassembly) I saw that check, and I copied it
if I was a wine dev and I read that yeah, you're not contributing lmao

>>108791591
>All users that daily drive Linux always have a partition with windows installed
I don't :)

>>108788514
>>108789998
>VALs
VLAs are not necessary or important. They are just a variable-siezd stack allocation. We already had a tool for that, it's called alloca(), and MSVC implements it just fine. I even defined this fancy macro to go along with it.

#define StackAlloc(Type, NumberOfElements) ((Type *) _alloca(sizeof(Type) * (NumberOfElements)))

>>108798209
>VLAs are not necessary or important

scratch space uniportant, got that mr. noob. as i said:

> nonono, no malloc babaloc substitutes, otherwise install..

> We already..

i dont speak wespeak