ReactOS is an open-source Operating System designed to be compatible with Windows NT. Version 0.2.8 sees the culmination of months of work since 0.2.7, and sees the project coming closer to the long-awaited 0.3.0 release.
ReactOS 0.2.8 Released
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
that this project is recieving an underwhelming amount of press and support. Once this project starts getting further underway, and further to completion, it will make some waves in the popular OS pool. Good work.
Writing reactos to be compatible with current windows drivers would be schweet
They’re all “HOORAY, we’ve reached 0.2.8… that’s closer to 0.3.0 than EVER BEFORE!!!”. And then the changelog starts off “Finalized implementation of the NDK and fixed all outstanding main header issues.” Hooray! I’ve been been so eager to get that thar NDK implemented for SOO LONG!
Ok, enough lame humour. I’m glad to see this product released, and it does take the project significantly down the road to possibly one day being useful.
Congrats ReactOS Team!
Why would someone bother with React OS … when you can pick up a copy of Windows NT on ebay for $40?
Because it’s open source..?
Maybe, but also because soon, there will be no more updates to the older winNT releases…
Why would someone bother with React OS … when you can pick up a copy of Windows NT on ebay for $40?
Yeah and what happens when Microsoft EOL’s NT4 (if it hasn’t already) and your business has to either (a) migrate systems or (b) hope Nothing Bad Happens ™ or (c) a market could sprout up to support legacy Windows installations because the source code is available.
Hmm, (c) does sound quite interesting.
I believe they have EOL’d NT 4. And are you serious about the rest of your comment? The vast majority of businesses will opt for (a). If computers are that important to their business they’ll be replacing machines as they age and the new machines will have a newer OS.
Don’t be so sure of that. I work for a place that was supporting OS/2 Warp 3 up until 2002 because of a home grown legacy app without a suitable replacement. If a viable continuation exists and is supported, it may be a more reasonable choice for some people than you think.
It’s all about a little thing called freedom.
It’s an interesting little OS, but I couldn’t see using it for serious work. What’s I’d really like to see is using this to improve windows executable support in linux, e.g. wine.
-bytecoder
“What’s I’d really like to see is using this to improve windows executable support in linux, e.g. wine.”
They have a very close working relationship with Wine. Just view their FAQ. 😉
//Hmm, (c) does sound quite interesting.//
It sounds interesting, but I doubt it could handle a big load of NT support.
Plus … geez, NT came out in 1996 or so. Upgrade already.
It’s not really an NT replacement, closer to a 2000 or XP replacement.
They have working USB, unlike NT4.0 and are working on DirectX (NT never really got above DX5)
They just mean an NT based OS such as NT, 2000 or XP. Who knows how much of NT is still in Vista, there are too many current applications to break compatability too badly. ReactOS may actually have a chance of catching up fairly well, especially once it becomes usable as an everyday OS.
Lots of NT, although refined and improved, is in Vista, since Vista is NT 6.0. Nothing wrong with that, the NT kernel is for the most part a very good design.
Who knows how much of NT is still in Vista, there are too many current applications to break compatability too badly.
Ever hear of virtualization? It’s sort of like backwards compatibility except you don’t need to go backwards to get it?
To be exact, NT 4.0 had DX3 build in (which, if memory serves me right, came with one of the service packs) but never had the driver-layer support required for DX5.0.
There was a beta software-only DX5 for NT, but AFAIR it was never officially released.
Cheers,
I believe they have EOL’d NT 4. And are you serious about the rest of your comment? The vast majority of businesses will opt for (a). If computers are that important to their business they’ll be replacing machines as they age and the new machines will have a newer OS.
What about the software that ran their business before? Who’s to say it will work on Windows Vista Server, or whatever is out at the time.
Plus … geez, NT came out in 1996 or so. Upgrade already.
I co-oped at a manufacturing plant that still ran core manufacturing processes on VAX bigger than 2 refrigerators from the 1980’s along with Windows NT4 servers. I’m sure it took many millions of dollars to develop those systems.
Like it or not, legacy systems abound. Nobody hates dealing with them more than me.
But if 0.3 is the long-awaited release I’m going to hold off trying ReactOS until version 0.3.1.
It would be interesting if reactos people support posix subsystem and allowing one to boot into posix environment directly. This will make a OSS *nix that supports binary windows driver. And may be adding a minimal wine like environment to redirect GDI/DX calls to X, to allow running win32 exe form within the posix subsystem.
This would sound like the dream OS for me…..
It’s not exactly what you describe but what’s wrong with cygwin?
Windows 2000 and Windows XP are just newer versions of Windows NT4. Will this React OS be compatible with all Windows NTs?
Will this React OS be compatible with all Windows NTs?
That’s the plan.
I can never access my XP partition (hda1) from ReactOS LiveCD. This means I can’t test it properly. Whenever I go to “My Computer” and click on C: drive, all I get is an empty folder. Any ideas? Looks like the volume isn’t mounted and if this is the case then how do I mount it?
I don’t believe it supports NTFS at the moment.
Ahhh that does make sense. Thanks for that! I remember now reading somewhere about ReactOS not supporting NTFS partitions.
I am a big FOSS fan, but ReactOS is simply not ready yet. Not ready for anything. ‘Nuff to say that upon installing both Win98 and ReactOS 0.2.7 in qemu, I can see React needs some 3-4 times more CPU power, is visibly much slower, and pretty much useless when it comes to any kinda work. Well, ok, solitaire works.
My 3c.
Don’t you think that comment was totally useless? We are talking about a 0.2.X release and most people consider a 1.0 release as >>not ready yet<<. In no way did the ReactOS people imply that 0.2.8 is ready nor that 0.3.0 or any other release short of 1.X will be. This is pre beta and will be for a while AND it’s not a secret.
I am a big FOSS fan, but ReactOS is simply not ready yet. Not ready for anything. ‘Nuff to say that upon installing both Win98 and ReactOS 0.2.7 in qemu, I can see React needs some 3-4 times more CPU power, is visibly much slower, and pretty much useless when it comes to any kinda work. Well, ok, solitaire works.
My 3c.
…..hence the version number 0.2.7….
What are the system requirements? RAM? Disk space?
Would be nice if they put that in their Wiki.
They list some drivers but the list is rather sparse.
It’s very interesting that, while I think MS is aware of this project, they did nothing to stop it.
We know from past experiences that Microsoft will not allow OSS developers to provide alternative implementations of their technology if they consider that strategic.
If we look at MONO experience, MS allowed them to create MONO but warned them from starting a new implementation of Avalon and Indigo even before they started, just as soon as they talked about that.
That means, IMO, they consider ReactOS to be a good thing or they would have warned such developers to stop.
A full-working ReactOS (compatible up to WindowsXP) could be good for them and could move those users who have lower needs to a Windows-compatible environment which is good because at least they wouldn’t move to a competitor system. Maybe they forecast a future where a slight part of their userbase could move to a free Windows-compatible system and that would be better for anti-trust reasons (those user would be formally not MS users) and would probably allow Microsoft to grow in the server field without worrying about anti-trust complains.
This is the same strategy IBM (and others) are using to sell their high-end OS (AIX) by introducing them a “compatible” or “similar” environment like Linux but MS controls its market.
I will keep an eye over this project because I think it has a potential, both for technical and pratical reasons.
A suggestion to developers: learn from others. Keep your eyes on XP (SP2) and resist the temptation to go for Vista technologies…
“A full-working ReactOS (compatible up to WindowsXP) could be good for them and could move those users who have lower needs to a Windows-compatible environment which is good because at least they wouldn’t move to a competitor system.”
Then why isn’t MS helping ReactOS. I think MS doesn’t like the project, why would they like something which could eventually take away users from Windows. Someone on an old version of Windows is still seeing the MS Windows splash screen every time it boots up, that results in brand name recognition which will come in handy should those people then decide to upgrade. There’s no benifit for Microsoft in this, their inaction so far is probably because:
– The project doesn’t have any significant funds that people know about, nor does it need such funds to continue existing since it’s OSS.
– If MS can’t destroy the project by getting them into an expensive and drawn out court battle then they actually have to find something these people are doing wrong before they can destroy them.
– ReactOS isn’t known to be in violation of any patents or copyrights.
– ReactOS is so small that any attempts by MS to interfere publicly with the project now would only increase interest in the project.
– Since it’s still only a 0.2.x release ReactOS has a very good excuse when people start complaining that it’s not “desktop ready”, so MS can’t win by telling people it’s a sub-par product.
Then why isn’t MS helping ReactOS.
MS wasn’t helping MONO too but they weren’t afraid about that project and did nothing to stop it. But as soon as someone started to mention Indigo, Microsoft made it clear to stay away.
However, I agree it’s too early to say something sure. However, don’t underestimate work ROS developers already did. Once you have the basis ROS now has, you can advance very quickly. The dirtiest job was to create the skeleton of the project to build the rest upon.
I agree that Windows system, by itself and contrary to other systems like Linux, it’s huge and has hundreds of services to implement.
At some point, I think ROS developers should choose what to refine and what to leave out.
I don’t think it can fully compatible with windows.It maybe a interest joke or a toy to play some little windows applications.,because windows’s underline data struct diffcultly be reversed
You would be surprised how much of the “private” data structures are published in the public debug symbols…
Which are, of course, perfectly legal to reverse engineere?
Cmon guys, ROS will be production-quality ready comparable to NT4 in 2010 or something, and by then it will be as amusing as dosemu is today.
Go write PaX for NT kernel, build a security framework that will hack into SRM and provide support for alternative security models, build an address space layout randomization driver, provide 100% functional windows support for linux x86 ELF executables (line!), build DTrace for NT kernel (something like windbg/kd scripting in user-mode, but integrate it for example into MSH namespace provider)…
There are SO MANY more productive and needed things today, and I’m afraid that you are focusing your talent on wrong things now.
I think you forget that most of the userland dlls are provided by wine, of course there is no fine line where reactos and wine overlap.
Also, reactos is opensource! I suspect they will rely heavily on the linux ntfs driver when writing the ntfs driver, and that isn’t the only example.
Which are, of course, perfectly legal to reverse engineere?
Depends on where you sit. In e.g. Slovakia or Czech Republic, it is legal to reverse-engineer a product for sake of co-operation with other products. IANAL, but as far as I understand, an EULA cannot take away your law-given rights here.
Which are, of course, perfectly legal to reverse engineere?
There is official ReactOS policy for this: http://www.reactos.org/xhtml/en/dev_legalreview.html
In addition, regarding to my previous post, the Windows debug symbols are public information and using them probably falls under the fair use (NOTE: I’m not a lawyer). The symbols are automatically downloaded by quite a few debuggers and disassemblers without asking for signing any kind of license.
Cmon guys, ROS will be production-quality ready comparable to NT4 in 2010 or something, and by then it will be as amusing as dosemu is today.
I tend to disagree on this one, but only time can tell. It depends on how many developers the project will attract.
Go write PaX for NT kernel, build a security framework that will hack into SRM and provide support for alternative security models, build an address space layout randomization driver, provide 100% functional windows support for linux x86 ELF executables (line!), build DTrace for NT kernel (something like windbg/kd scripting in user-mode, but integrate it for example into MSH namespace provider)…
I consider some of these as a good ideas (and some not so good), but the main point is that this is open source software and people are free to choose what do they devote their spare time to. Moreover it can help people to learn and understand the NT architecture, which you would need solid understanding of to accomplish any of the projects you mentioned above (actually most of them would need hacks to be injected into the real NT kernel; with ReactOS you have the source and you can implement them without resorting to hooking and other “unsafe” practices)…
Edited 2005-10-31 14:06
Which are, of course, perfectly legal to reverse engineere?
As others said, it mostly depends on where you live. Reverse engineering is perfectly legal in the whole EU. So this is not a problem for (soon-to-be) 400millions EU citizens.
There are SO MANY more productive and needed things today, and I’m afraid that you are focusing your talent on wrong things now.
There are people (like me) who think there’s room for a project like this. There are people who are not satisfied with evolution (or lack of) surrounding projects like Linux and who think that a Windows clone could find a good way among users.
If ROS was a completely new OS, maybe you could be right (loot at SkyOS: nice stuff but being a completely new OS IMO is a huge disadvantage…). But if (big “if”) ROS can get to achieve good Windows NT/XP compatibility, that would be something great.
Let’s discuss the really important stuff, how *does* the BSOD screen look in ReactOS?
Edited 2005-10-31 09:45
Reguardless of the legal issues reguarding reverse engineering, ReactOS is not a reverse engineering project. It is a re-implementation of the windows kernelspace API. (Just as WINE is a re-implementation of the userspace API)
The API is public. Without access to the API, developers would have no way to write apps that run on windows. There is no reverse engineering required to re-implement the API.
It’s like the stearing wheel, gas pedal and brake pedal in a car. A car can be gas, desiel, a rotary engine, or whatever. It doesn’t matter. You drive them the same way even though the engines are very different. ReactOS is a very different engine, with the same stearing wheel and gas/brake pedal conventions that windows programs need in order to drive the car.
What I think alot of people don’t understand is the positives of ReactOS and its ability to test the underlying API of Wine; not only are people compiling the wine dlls and replacing components on Windows with the opensource equivilants, like what is being done with Haiku-os, but also tested on a stand alone environment.
For me, I’m actually rather excited about the news that the Solaris patches are finally being accepted into the main wine tree; hopefully once the Linux compatibility becomes production ready, coupled with a stable 0.9 of Wine on Solaris, the number of possibilities for UNIX on the desktop will increase.
Reminds me of the old KDE statement of, “Is the desktop ready for UNIX?”
it is a great idea and i wish to project good luck ang getting f–king better!