Ok, somehow I went off on this tangent in a conversation just now. What would happen if the linux/open-source community worked on cracking the windows kernel? Create an open-source windows kernel replacement. Just the kernel mind you. It would still run the windows shell. This would bring a bond between normal users/gamers and the open source community. And linux programmers might even get ideas for ways to support windows games within linux in a way not yet seen. Another possibility: Later, the new kernel is released with an open-source shell, such as litestep. Huge advancement: The average home user isn't capable/isn't willing to learn linux inside-out just to be able to effeciently use it on a daily basis. This would let the user keep their beloved interface yet keep on top of security with a kernel that is maintained by the open-source community. In short: windows + open source = security+games+ease of use Just a random idea...
Wine TransGaming Codeweavers Actually, I believe if you reverse engineered the Windows kernel it would be illegal. The above sites are attempting to recreate the windows kernel in a Linux enviornment without hacking the actual kernel or dll... I've successfully installed and run Internet Explorer 5.5, Office 2000 and a few other apps natively in Linux using Wine. Problem is that it's quite hit and miss for other apps - especially games.
as far as I've heard/seen wine is quite slow for games I have yet to get winex running and crossover costs $$ And they're not really re-creating the windows kernel as much as emulating it... getting it to work with the linux kernel. This would be a seperate entity. As for copywrite issues. I don't see how it's different from shell replacements such as litestep?
Wine is a bit slow with DirectX and stuff like that, but the price is right and they are heading in the right direction. Just a couple of years ago it was a serious dog, but they've streamlined the code and it's been running faster and faster. The best thing to do is to copy ALL of Windows DLL's and use them instead of the Wine native ones. I do that and Wine is faster because of it. I've never screwed around with WineX, but from what I heard it is a pain to get configured, but when you do, the DirectX stuff is MUCH, MUCH better than regular Wine. Yea, Crossover does cost money, but it is by far the best implementation of the Linux/Windows enviornment I've ever used. I got a free copy from the source to test the feasability of migrating workstations over to Linux and it runs real well. I'm not sure if I read your initial post correctly now. Are you talking about doing something like Mac OS X? A Windows GUI onto of a Linux kernel or a complete Windows / Linux integration by hacking the Windows kernel? If you're refering to the Mac OS X thing, remember, there is KDE. Very similar in the feel plus you can get skins so it looks identical to Windows 98 and 2k. If a Windows / Linux integration, Microsoft would never allow that to happen. If you released something like that, you would have to have the source for Windows (or hack it) to fully integrate the 2 OS's which would violate the DCMA among other things. Billy would be all over you for that!
How do you think Windows integration has occurred in Linux (NTFS support, etc.), people get WIndows, crack it open, and figure how it works, and then build it in the Linux environment. However crackign the Windows kernel in a way you describe would create somethign akin to Lindows, only far far worse. If that's possible. Also Windows doesn't ahve a kernel like Linux does, so then you'd have ot compensate for that, and the two really work very differently, and building a sort of "super OS' like that would not work. At least well.
Very true indeed. However, I'm under the impression that he's refering to the Windows API as the kernel. I know that Wine et al is trying to implement the Windows API's in a native Linux enviornment, thus essentially cracking Windows, but if they want to conform with the DCMA, they can't implement any Windows code. It has to be re-written. I also agree that it will never work as well as a native Windows enviornment would given the differences, but with processor, memory and fsb speeds going they way they are, who will care if Dreamweaver runs 20% slower on Linux than Windows? It'll still run faster on my xp2100 with Linux than my old p3 600 with Windows! What's wrong with a super os!? The stability & security of Linux with the software library of Windows? I've never used Lindows before, but from what I gather and read, it's pretty much just essentially RedHat with a custom Wine distribution.
My thought was this: Two main parts to windows- shell and kernel Replace the kernel but have it use the windows shell. It really was a random thought, but it was one I didn't hate. The idea isn't a "super-shell" at least not in my mind. At first it would just be a "haha, lookee what I can do" kinda thing. But if the show-off kernel were open source, people would start going, "If I change this, that will run faster" or "If I add this bit, that security hole won't exist" Pure random thought. I don't know much about the inner-workings of windows, but hopefully, if this is impossible, the reasoning will teach me more.
No I'm afraid it's solid evil. No normal creature would be able to construct such a vile thing without being killed in the process. Hence, failing it's creation. Unfortunately it didn't fail. Also the bit about not worrying about performance on fast systems, well what about the latest and greatest game that barely plays on the top spec system at the time, then a Linux variant would run worse, so getting a fast stable running system with this Linux/Windows hybrid would be required. Although it really doesn't matter, as if your going to do that much work to get the best of both worlds, you might as well just build your own OS from scratch, instead of trying to rebuild Windows and Linux (one of which is closed source), since it would be easier. But it doesn't really matter since we went way to far indepth for the scope of the original idea anyways
But by cracking and building replacement kernels would lead to higher-grade linux/windows/(something else?) hybrids, bringing the best things together into one package and leaving the crap in the dust... at least the way I see it in my head...
Well, having Windows stuff work under Linux would be very sweet. That would convert me pronto. But, I would want EVERYTHING to work FLAWLESSLY before I threw out the XP cd. That means games as well. I think, though, that as soon as you get something that runs Windows programs well under Linux, Billey Gates is going to start suing the socks off you before you can say "penguin" Wine is a step in the right direction, but they have a ways to go yet
Between a kernel replacement for windows and an emulation of windows in linux, you'd have two huge steps in the right direction.
