The last ReactOS release is already twp years old, and there seemingly hasn’t been any news since. Of course, the project has not stalled, and in a newsletter the project details the progress that’s been made since 2021.
In the last year and during the beginning of 2023, the ReactOS developers and contributors alike are working on many parts of the project, the top focused area being the kernel. Other areas that aren’t kernel-related are the applications, specifically our Paint and Notepad programs, the Input Method Editor (IME) as well as other stuff such as the ReactOS testing infrastructure.
There’s steady progress on the x86-64 port, improvements to the Security Subsystem, and more.
The longer I am away from Windows, the less I care about ReactOS. Back in the early 2000s and maybe up to 2010-2015 I had some interest in ReactOS, but I haven’t used Windows itself since 2004-2006, and I couldn’t be happier. I actually forgot ReactOS even existed, and to be honest… I’m not sure I see a point in it anymore. Windows as a target keeps moving, even breaking itself from version to version, and ReactOS is still unstable, incomplete alpha. And it’s chasing a desktop that lost relevance in my mind many years ago. Meanwhile Linux itself is getting so much better in so many ways, and it is amazing how for it’s come in the last two decades or so since I’ve been experimenting, using and tracking it.
I’m at the point where I literally only see Windows as being useful for gaming and for updating the firmware on radios and wireless game controllers… and these days, Linux is even creeping up on gaming. I’m not so sure I see where ReactOS fits in if the OS it’s trying to replicate is so pointless these days.
UltraZelda64,
There are a lot of legacy and industrial systems that depend on Windows NT.
Same with DOS, which has been happily replaced by FreeDOS in recent times. Many BIOS update and similar “boot disks” carry FreeDOS in them. Also if you are into emulation (DOSBox, pcem), their command line tools would be very useful.
Yes, ReactOS is still “not there yet”, but they have made enough progress to become useful with some effort.
ReactOS is still significantly incomplete. It has good promise, but it seems that 90% compatibility is still “10 years away”, much like it was 10 years ago.
I work on industrial systems like that.
Anyone still relying on XP era NT or earlier is being negligent. We usually convert such systems to PLCs and call it a day. Sometimes they also have an accompanying windows app but those get upgraded over time and don’t rely on vintage hardware or software.
Tell that to all the big banks in the US (no really, please tell them!) who still run OS/2 on their ATMs. I get that “if it ain’t broke, don’t fix it” is a thing, but there comes a time when you should upgrade to a somewhat modern back end, if for nothing else than sanity’s sake.
Morgan, cb88,
For a “non software” company, IT and development are usually costs with no clear connection to revenue.
If they can still sell those machines, and have no issues with any certifications, they would unfortunately have zero incentives to upgrade any of the software.
If and when software becomes a competitive issue, or they are required by external standards, only then they would want to upgrade.
Remember, they are in it to sell the ATM, not the software.
FreeDOS is great. It’s stable, compatible and mature. ReactOS is… none of those, except for maybe compatible. But Wine is compatible too, and the operating systems it tends to be run are both of those other things. Haiku for years has felt far more complete and usable to me than ReactOS ever has. Every time I give ReactOS a try it’s unstable to the point of unusable, and much worse than even Windows Me way back in its day.
The “FreeDOS as a BIOS update host” point is still interesting though. ReactOS is a long way from being a user’s primary OS – but it’s not that far from being able to host a single dedicated program, particularly if the program can be modified to run on it.
Yes, what could be more important is that ReactOS also supports Windows NT drivers as well:
https://reactos.org/wiki/ReactOS_FAQ#Is_ReactOS_able_to_run_Windows_software_and_Windows_drivers.3F
For many applications, “just running it on Wine” might not be possible if they come with a proprietary hardware with no open source alternative.
As somebody that does not contribute to them at all, ReactOS has to be one of the most frustrating projects for me in terms of how it is managed. At the stage it is at, why hold back releases literally for years? The communication has been consistently terrible forever. The project is positively hostile to its community. The forums are bursting at the seems with “Stop bothering us. Adults are doing things you could not possibly understand. If you are not going to help, go away.” though they are mostly abandoned with the communication that does happen moved to Mattermost ( good luck figuring that out from the website ). The end result is that progress is glacial.
They are doing hard things in their spare time I understand. But harder than what other projects are tackling? I am skeptical, especially given that absolutely massive advantage that they get from WINE.
I cannot help but compare their approach to something like SerenityOS with its “Hello friends” founder, engaging YouTube videos, detailed monthly updates, frequent interviews and conference presentations, and developer “office hours”. SerenityOS has made just enormous progress in a very short period of time. Starting with a single developer, they have built a complete OS from scratch along with an impressive app ecosystem. ReactOS is touting improvements to their Paint program. PixelPaint in SerenityOS has layers and supports an array of image formats ( all written 100% within the project ). While ReactOS does not want to distract themselves even getting a browser working, the SerenityOS folks have created an entire web browser from the ground up ( including support for HTML, CSS, JavaScript, SVG, audio, and video ).
That said, despite my pessimistic description, I am an enthusiastic supporter of the ReactOS mission. There is an absolutely immense number of Windows-only applications. Having an Open Source home for this treasure trove is important. Not only that but, increasingly, much of this software can no longer be run on current Windows. The same is true for the enormous universe of hardware that has only Windows drivers. Those that rely on these things need ReactOS. With so many people and companies that would benefit from a ReactOS that was further along, it is almost a miracle that they have succeeded in attracting so little support and participation.
Windows does add to their API but the pace of change outside of GUI frameworks is not really that blistering. The API is quite stable generally and WINE keeps up with a lot of what does change. It is interesting that you say “Linux is creeping up in gaming” as I would say that is mostly happening because WINE / Proton are supporting the Win32 API on Linux.
It is exciting to see ReactOS implementing 64 bit as this is essential for running much of what end-users will want to run these days ( a somewhat current web browser as an example ). Even with the project as it is, I wish them luck and look forward to their success.
ReactOS is exactly the kind of project I would love to contribute to if only the project were a bit friendlier.
I’ve never experienced any of this on their forums, so I’m guessing this is just you imagining it. Most of their developers don’t even use their forums, and most posts there are from teenagers and total noobs to coding, so the responses aren’t great and usually similar, but they aren’t the condescending replies you make them out to be. Also you seem to think they have tons of coders that can accomplish these tasks. They aren’t holding back releases, they’re just releasing less to spend more time on actual progress rather than fixing regressions. usually they only have 20-30 developers working on it part time, so you’ve misinterpreted that they don’t have the resources, and if you want to talk to them and get working on it you head to Mattermost (IRC replacement), not the forums. Also it’s OS development in C, so far less hand holding is expected and a lot more skill is required to start out.
Its not you, it’s me. The project is not hostile or condescending. The problem is all the “teenagers and total noobs”. Got it.
It is not worth getting into a back and forth. My comments were made in the hopes that somebody might hear them understand what the current approach is costing the project. If I am wrong or my comments are not helpful, just ignore me.
Perhaps I will take another run at one of your points though as it I see it as one of the biggest barriers the project erects to prevent greater community participation.
“it’s OS development in C, so far less hand holding is expected and a lot more skill is required to start out.”.
Again, I will mention SerenityOS which is written in C++. Most people consider C++ harder than C. In that project, the project founder has attracted programmers that initially knew only Python or JavaScript and learned C++ just to participate. One of those, Linus Groh ended up becoming a co-maintainer of the project and has become such an expert that he presented their “from scratch” JavaScript engine to the standards body that stewards the language specification. I am not sure how old Linus Groh is but he looks pretty young in the videos. He may have been pretty close to a “teenager and total noob” when he started with SerenityOS.
As a result of how the project is run, SerenityOS now has hundreds of part-time contributors and enough financial sponsorship that some of them have been able to go full time. The idea of an Open Source Windows is inherently attractive and valuable to many people. It remains my opinion that ReactOS could have an absolutely massive and vital dev community if it was managed differently. In a different timeline, the Steam Deck is running ReactOS and ReactOS has more desktop users than Linux.
In my view, SerenityOS has attracted so much attention and participation because of the “Hello friends” and “I hope you learned something in this video” approach taken by the founder and the pragmatic accountable, open, and respectful culture he has built-up around the project.
You say that I am imagining the cultural problems in ReactOS and that my thesis that a more “Hello friends” attitude would help it to perform and grow is misplaced. Maybe.
I am not sure if @dark2 here is @dark on the ReactOS forums but I just looked at the ReactOS forums and here are a couple of snippets that @dark left on those forums just today:
“your legal advice is worthless”
“Honestly I’ve only skimmed his response because it’s just not worth it”
Anyway, regardless of if I am right or wrong, I wish nothing but enjoyment and success to ReactOS, its devs, and its fans. I am very thankful they are working on it and I would very much love to use it one day. I have considered contributing to the project many times. Who knows? Maybe someday I will.
I have been a happy Linux desktop user since transitioning from OS/2 to Linux in the mid-90’s ( when it did not even make sense ). That said, I am not sure I am rooting for an OS mono-culture.
Windows has an interesting architecture and is one of the few relevant options that is not just a UNIX-a-like. It shares a degree of heritage with VMS ( through Dave Cutler ).
Again, though I like Linux, there is an argument to be made that perhaps a desktop should be a bit more opinionated ( and standardized ). Perhaps that is one reason desktop Linux has struggled. A world where Linux dominated IoT, the server, and the cloud but ReactOS / Windows ruled the desktop would be ok with me. Who knows, a descendant of ReactOS may someday embrace and extend the desktop away from Microsoft. At the very least, it can provide an alternative to desktop fans as Microsoft guides Windows away from its role as a traditional OS to being more of a set of services that you subscribe to.
There are other desktop options of course. Haiku seems to be getting close to being a sufficient daily driver for at least some people. I am encouraged by the momentum in SerenityOS. RavynOS has a long way to go. That said, I do not really believe that either of these is going to topple Windows on the desktop. As long as Windows dominates the desktop, we should be glad for ReactOS.
I’ve claimed for years that the reason for Linux’s failure on the desktop is standardisation, or lack thereof.
Google proved that Linux is ultimately a very stable platform to build upon, with their development of the Android userland. This provided a standardised, open source platform for mobile devices, that not only ships on the majority of smartphones, but was largely the reason for the discontinuation of Windows Mobile and other competitors like Symbian, BBOS, PalmOS etc.
Again, for the reasons of consistency, i believe that the most promising FOSS OS for “desktop dominancy” is Haiku. With a stable API, ABI, and surprising amount of hardware, support Haiku is a stable and very usable OS. On top of that, the suite of software available for it has grown exponentially in the last few years, with most any task you can do on Linux, having a ported, or home-grown alternative available on Haiku. The chicken-and-egg problem is largely moot now. However, though it is crisp and clean, the biggest letdown of Haiku is the UI. Though i don’t think it would take much to bring it into the 21st century, a significant (but sympathetic) redesign, maybe along the Win11/MacOS UI lines, would definitely improve Haiku’s visual appeal.
Sadly, I don’t see Haiku evolving beyond where it is now. There are patches getting fixed but almost zero native application development. The contributing community is very small at this point.
I don’t particularly see Haiku ever getting past being a nostalgia project, and the interface doesn’t help. It will likely end up like SkyOS, unable to keep up with new hardware drivers, but it will stick around for eternity due to being open source. I guessed the first example would be good desktop scaling for high DPI screens, Googled it, and confirmed it doesn’t have a consistent method of doing that without tweaking individual elements and restarting. Another technically works on the surface, but just a little digging reveals the underlying support for modern stuff isn’t there.
As long as ReactOS doesn’t ditch 32bit. I think it is a great (potentially) alternative to running on Retro gaming machines. Screw fiddling with windows 98…
There’s a lot of software for 98 that just won’t run on NT. ReactOS definitely isn’t a replacement for 9x.
However, given the simplicity of early DOS based Windows, I’m very surprised there’s never been a serious attempt at making a 3.x/9x FOSS clone.
There was a project called Freedows way back in the day that wanted to be a Windows 9.x clone. Creating a DOS extender that implemented the API may not even be a super hard task given the tools we have today. What would likely be an absolute royal pain would be driver compatibility.
I’ve been looking at “HX DOS Extender”, seems a lot of the work to get Windows 9x support might already be there. All it needs is a window manager and some glue logic to stitch WINE and ReactOS code to implement the rest of the userland.
Driver support and the kernel side might prove to be a bit trickier, but it depends how far you want to go down the compatibility rabbit hole.
Pretty sure that was the project that became ReactOS. Once it became apparent that 9x was a technological dead-end, the devs changed tack and focussed on an NT clone. Freedows essentially died at birth.
I was going to say that Freedows became ReactOS but I could not remember for sure and a quick Google failed to turn up confirmation. Thank you for flagging this.
Most of the 98 software already runs in WINE, it’s very few applications that don’t run in WINE now. There’s some DirectX 5 API stuff I know WINE hasn’t implemented (look into Swapchains in Direct3D/DirectX1/2/3/4/5 era that’s still missing). WINE also has a problem with 24-bit/32-bit colour (Intel/AMD Mesa drivers only support 32-bit colour and 24-bit renders as a black screen) You can usually work around colour depth problems by running X11 in the correct colour depth. Other than those things the only other issue I’m aware of is WINE and running custom VXDs, obviously WINE isn’t build for handling Windows 9X hardware drivers.
I highly doubt anything will save you from fiddling with software from the 95/98 era in the first place. The bugs are even in the hardware itself.