Linked by Smith Johnson on Thu 27th Sep 2007 15:22 UTC
Windows According to at least one blogger, Microsoft should abandon Vista before it's too late. It would appear he's not alone in this opinion, as Microsoft has begun allowing users to downgrade back to XP. Amongst the reasons? Poor sales figures and shoddy Vista "Extras".
Thread beginning with comment 274902
To view parent comment, click here.
To read all comments associated with this story, please click here.
RE[2]: A sidenote...
by raver31 on Thu 27th Sep 2007 19:51 UTC in reply to "RE: A sidenote..."
Member since:

Yeah the NT kernel is pretty robust, it does not need to be Unix like.

Reply Parent Score: 2

RE[3]: A sidenote...
by SReilly on Thu 27th Sep 2007 20:13 in reply to "RE[2]: A sidenote..."
SReilly Member since:

Yeah the NT kernel is pretty robust, it does not need to be Unix like.

I think that both kernels share many of the same features. NT is a very modern, robust and secure kernel architecture. Arguably, the userland of both NT and Linux are very different but all in all, I don't think that the NT kernel needs all that much work.

I think that, if MS does want to make a break with the past, a good reason would be the discontinuation of the NT kernel in favor of Singularity. That way, NT could be virtualized for backwards compatibility and Windows could move to another level altogether without needing to worry so much with running old software.

Just my €0.02 ;-)

Reply Parent Score: 3

RE[4]: A sidenote...
by PlatformAgnostic on Thu 27th Sep 2007 21:59 in reply to "RE[3]: A sidenote..."
PlatformAgnostic Member since:

There's a slightly better approach. Singularity can run alongside NT as a special kernel-level process. The main NT scheduler would schedule the threads, but the singularity runtime could execute in a kernel code segement and allow direct calls into the kernel entrypoints rather than the usual hardware-protected system call mechanism.

I don't know how long it will be before the Singularity technology makes it out of Research. I read an interesting paper by the directors of the project and they seemed to be saying that the research codebase will never go directly into a product.

If a Singularity-style product goes into the market, though, it looks like it will be hard for anyone to compete. It doesn't take a genius to predict that MSR is patenting all of the core technologies needed to make that work, especially the code verifiers and the theorem-proving compiler.

Reply Parent Score: 2