n3rrd

@jgp
If you need "as much as you can get", why are you sticking with a rather modest 4GB?

@grahamj

How about we relax with the conjecture? Manipulating the tab interface to work more like classic windows on screens with a reasonable amount of real-estate is possible.

Additionally, multitasking doesn't necessarily involve having two windows open side by side. Many smart-phones multitask even though their screen real-estate doesn't allow for windows. All it means is that a process can continue doing it's thing in the background without having the app pause indefinitely or shut down.

@jonnythan
Well, at least the flexibility of a smart phone isn't wasted on you...

@iErik Thank you. Someone who isn't an idiot.

No, the keyboard operates on magic.

Jack,

What does the availability of Snow Leopard matter? This isn't a product aimed for shelves. Without much research into what exactly GCD is, and how it differs from other methods of shoe-horning multicore support into standard kernel/OS design, I have difficulty believing the scope and effect of this technology is as extensive as that in Microsoft's research.

I don't give a damn one way or the other but I don't see why money would be invested into researching a topic understood well enough that it's already been implemented in main-stream consumer level OSes by other competing companies.

This is so ridiculous. It isn't just "multicore optimized", it's multicore specific. It's a total shift in approach (from standard consumer OSes) to OS design. Fuck off with the Apple, Microsoft, *nix bullshit. This is an OPEN SOURCE project. The results of the research will benefit OS design, and while there will probably be components not released that are specific to Microsoft development, it will help everyone step up their game.

I'm not completely familiar with everything being discussed in the PDFs on the Barrelfish site, and a lot of it is over my head, but if you'd pull your head out of your asses and take a peak, do some reading and research, you'd really see what this project is about. I've also read that a lot of similar research was done in the 80s, but the game has changed quite significantly and hardware is drastically superior to that which existed back then. It's very possible that a solution that wasn't viable on previous hardware is viable now.

Sometimes I find it difficult to swallow the stupidity of some people. Jesus christ.

wickedphoenix, I'm not sure that this has anything to do with general system RAM, but instead refers to shared system cache, and caches on other cores (L2/L3). I may be mistaken, though. I also may have read your post incorrectly :)

I agree when talking about CrossFire or SLI; but with the ability to purchase any new, reasonably priced card, toss it in the system, and have your performance increase is very useful.

Very Gameboy Colour.