Singularity Possible Still?


Recommended Posts

Since Windows Vista, Microsoft has been separating different areas of Windows this has continued with Windows 7, Windows 8 along with their Server Line. Each area has grown isolated as in nowhere near as dependent on other parts of the OS. Do you think we could be headed to removal of the NT kernel (except for virtualization) and moving to a more developed version of Singularity which was developed by Microsoft Research?

More info on codeplex.com

http://singularity.codeplex.com/

Link to comment
Share on other sites

Since Windows Vista, Microsoft has been separating different areas of Windows this has continued with Windows 7, Windows 8 along with their Server Line. Each area has grown isolated as in nowhere near as dependent on other parts of the OS. Do you think we could be headed to removal of the NT kernel (except for virtualization) and moving to a more developed version of Singularity which was developed by Microsoft Research?

More info on codeplex.com

http://singularity.codeplex.com/

Unfortunately, not as long as the backward-compatibility boat-anchor hangs around the neck of Windows as a platform.

 

For all too many developers (let alone users) backward-compatibility means all the way back to XP - resistance to anything that isn't backward-compatible to XP, and especially in enterprises, has gotten more and more vociferous.

  • Like 1
Link to comment
Share on other sites

For all too many developers (let alone users) backward-compatibility means all the way back to XP

XP? I still occasionally play a few games from the 95 era ;)
Link to comment
Share on other sites

Comments from the link:

 

Microsoft, when are you going to support another processor instead of x86 every time. No wonder devs are switching to Linux and away from managed code development. Being able to build this for an ARM 6 or 7 core would make this project actually worthwhile, instead it's a mess just the same as .net micro framework!! The only support you can get for .netmf is to pay someone stupid amounts of cash!!!
by DanRiches on May 23, 2013 at 6:06 AM
     
????????????? 
by yaucyj on Jan 14, 2013 at 1:37 AM
     
I think that any continued development and "out of the box" thinking (Windows, UNIX, etc,) is good and can only improve future OS development.
by DanteLA on Nov 26, 2010 at 5:58 PM
     
Too bad I dont see any development on it. 10-26-2010. would be nice for small systems that don't need windows. If you like this idea, please check out Cosmos. Just google it, er i mean Bing it.
by stratemeyerjw on Oct 26, 2010 at 1:45 PM
     
I think this is Lame. All Microsoft is doing is an another opensource kernel & I doubt the sincerity of Microsoft. The fact it's written completely in C# and runs on Microsoft's proprietary CIL is no more an advantage then just utilizing the proprietary .Net platform. In essence this could be done with Windows, just that Microsoft wants to make a good show. Rember MS Cairo & Windows 2k/XP source ?
by barterpc on Jun 17, 2010 at 12:30 AM
     
Jeszcze nie wiem czy podo?am temu systemowi wi?c zaczynam szuka? bo sam Windows ju? mnie m?czy i ci?gle udoskonala? go cza bo ma dziury szkoda i? nie wiele jest napisane o tym systemie 
by bronekb on Dec 13, 2009 at 11:46 AM
     
It is great initiative to build a managed OS> But it needs TFTP serevrs which means it cannot be used by everyone on Virtual PC and it is not that easily booted up.
by Borgdylan on Nov 29, 2009 at 3:49 AM
     
I really like this project.Have fun.
by nbby on Nov 6, 2009 at 12:04 AM
     
Is the project staing alive? I don't see any new release from 2007 and we are now in 2009 :(
by AresSieran on Oct 10, 2009 at 5:52 PM
     
We must evolve the OS as well of hardware.
by inhesm01 on Sep 22, 2009 at 7:03 AM
     
a fundamentally new approach to operating systems
by idamlaj on Apr 15, 2009 at 6:22 PM
     
Very cool project... i think is very good for those who want to see how an OS is made of from C# :) GJ 
by TheKimi on Mar 26, 2009 at 1:31 PM
     
this is so important to me
by dedunumax on Mar 17, 2009 at 12:38 AM
     
I just stumbled upon this release of Singularity. Thus far, I have liked what I have seen. The things that I like about this project is that it is well documented and offers a chance for new advances in Operating System design.
by eight190 on Jan 19, 2009 at 2:48 PM
     
most awaited
by nikopol on Nov 14, 2008 at 7:33 AM
Link to comment
Share on other sites

Do you think we could be headed to removal of the NT kernel (except for virtualization) and moving to a more developed version of Singularity which was developed by Microsoft Research?

Perhaps you are thinking of Microsoft's Midori?

Link to comment
Share on other sites

Short answer, no. Longer answer, no, there's no point.

Singularity was a research experiment first and foremost, NT does everything they need and isn't in need of replacing (Considering 99% of what makes Windows "Windows" is in a layer above the kernel, all switching the kernel would do is break drivers and some apps)

Link to comment
Share on other sites

When we're talking about something like a kernel, the only thing .NET would really offer is memory safety, you don't have the sandbox or anything like normal apps.

And what would be the benefit of running different kernels for different apps? (Calling into native code from managed code is fast, it's the opposite that's slower)

Link to comment
Share on other sites

That's pretty much what I said, memory safety is "code safety", that stops stuff like buffer overflows and prevents stuff like use after free (Which they'd have disabled in the kernel anyway, that stuff is slow), when we're talking about a desktop app you've got stuff like sandboxing thrown in, which wouldn't be involved with a kernel.

Also, 99% or so of the OS you're running is code that lives outside of your kernel, using a managed kernel wouldn't stop exploits in your shell, etc. (Could even make it easier, MSIL is quite easy to decompile to investigate)

Link to comment
Share on other sites

This topic is now closed to further replies.