Everyone loves a killer feature: that must-have capability or technology that prompts you to plunk down your hard-earned cash in an effort to upgrade your computing experience.
In the case of Windows, there have been precious few versions that included a truly killer feature. Windows 3.1 was a killer version because it allowed PCs to finally break (or at least reduce the impact of) the dreaded 640K barrier. Windows NT was a killer version (at least for power users) because it introduced the concepts of client/server security and true, hardware-based memory protection to the environment.
[ How to choose between 32-bit Windows 7 and 64-bit Windows 7. | Get InfoWorld's 21-page hands-on look at the new version of Windows, from InfoWorld’s editors and contributors. | Find out what's new, what's wrong, and what's good about Windows 7 in InfoWorld's "Windows 7: The essential guide." ]
Windows XP was a killer version because it bridged the gap between the consumer (Windows 9x) and business (Windows NT) computing spaces. And though generally considered a flop, Vista was a killer version in that it forced the Windows ecosystem to evolve beyond the Windows XP paradigm and thus paved the way for Windows 7.
Which brings me to my main point: Windows 7 is a killer version -- but not for the reasons you think. It's not because it fixes Vista's many faults -- it doesn't. Rather, it glosses them over with fresh paint and behavioral tricks.
It's also not because of the new UI. Although I'm a huge fan of the new task-bar-driven interface, much of the underlying concept is merely a rip-off of the Mac's aging dock metaphor. And it's not because Windows 7 is somehow lighter than Vista -- testing shows it takes up about the same amount of RAM when executing an identical workload.
Get the independent advice and expertise you need to support a virtual workforce.
The increase in Linux popularity has increased the frequency and sophistication of malware attacks. Read this 2 page white paper now to learn how you can protect your Linux environment with real-time protection that is certified by all major Linux vendors.
Download now »Ensuring acceptable application delivery will become even more difficult over the next few years. As a result, IT organizations need to ensure that the approach that they take to resolving the current application delivery challenges can scale to support the emerging challenges. This handbook elaborates on the key tasks associated with planning, optimization, management and control and provides decision criteria to help IT organizations choose appropriate solutions.
Download now »A common misconception is that mid-range storage requirements are dramatically different than that of a larger enterprise. Mid-range storage users may require less capacity, but they have similar functionality and management requirements. This ESG paper examines mid-range storage needs and reviews a new solution that adjusts size while retaining value, performance and functionality.
Download now »Read that again more carefully. He said Win7 is 250% faster than XP on the same machine, not that Win7 is 250% faster on an 8-core than a 1-core....
Oh no no no, dear child. As much as it pains me to have to say it, RCK has this one nailed. Thread management IS the killer improvement in VISTA SP3 (I'm sorry W7 is a sham). With NO Applications written as parallel applications and application domains that resist parallization, 250% improvement is phenomenal! Your faith in the Unix stack is misplaced, with the all IO is a stream paradigm, Unix has exactly the same issues. Massive scalability on parallel problems comes with clustering not SMP.
Perhaps I said poorly what I was trying to get across, but when I read what you wrote, it sounded to me, like you agreed with what I meant. You are dead on about different solutions fitting different problems. I just have not found a class of problems Windows fits well, but I am impressed with the improvement in performance VSP3 provides on a mixed load of Windows applications. Human bounded interactive environments are challenging to fit into a parallel paradigm.
I was also gratified you mentioned Solaris. I agree it IS the standard with which to compare 'NIXs. Not only does it's clustering and SMP stand out, but it's virtualization features and amazing files systems support. Sun fanboy I am. But I maintain, that UNIX as a specification, has a fundamental flaw in the "all IO is a stream" concept. This is the point. Parallel processing, at its most efficient, will have parallel IO. Like you said, fit the solution to the problem.
Servers are another story. It's a good thing that Vista/7 shares a lot with Server 2008, because that's where multi-core and lots of memory will pay off. Any extraordinary needs a user might have, can be performed on a server in the cloud. That is, assuming a decent network connection, right Regaug?
I bought a machine with Windows ME on it and tried to go back to W2000. Turns out the registry had changed with ME, and my box would not boot. So I bought XP and installed that. The box worked - I could boot up - but very slowly. I had to clear everything and start over. I decided there and then that I would try very hard not to ever buy another OS from Microsoft again, especially at their highway robbery prices.
Subsequently bought an iMac and wish I had gotten a Mac Book instead.
I haven't seen anything in any releases since XP that would make me want to upgrade. I can see that businesses would need to upgrade for support purposes, but, as an individual, I don't care to pay again for the "MS-Windows premium" on any computer I buy.

Sign up to receive InfoWorld Resource Alerts

1 reply