OSDev.org

Seriously, the only real changes apart from hardware support etc. etc. between Windows 3 and Windows 7 (I can't compare Windows 8 because I don't have it; Mom uses Windows 7) is that the interface is more bloated. I don't know what people mean when they say newer Windows is faster. The processor may be faster, but I get typing sooner in Windows 3 because I don't have to spend half a second waiting for the window to fade into view first. And to be honest with you, I have found Windows 3 to be very stable and not have any noticeably worse performance when carrying out everyday tasks. Web browsing may be the exception, but I can tell you that I have an internet connection on my Windows 3.1 computer and have posted to forums from it before.

When I talk about bugs in Windows I mean real bugs: I don't mean a bugs which crash Ubuntu copying an exceptionally large file (blame the user for pushing it so hard); I mean a bugs which corrupt the registry and which have been there since Windows 95. My Windows XP system is currently crumbling as I write for no good reason. It's simply a case of that one thing after another goes inexplicably wrong - I mean silly things like explorer crashing when I right-click on an item in the start menu. It's been getting worse and worse over a few years. A Linux system would never progressively deteriorate.

Mom's sick of me making that joke.

A Windows system won't progressively deteriorate either. You probably have broken software, or maybe even malware. I can assure you I've never seen Windows XP do something like that without the root cause being traced to some piece of software that had been installed.

Or registry fragmentation from a long history of installation, uninstallation, and configuration. That will at the very least progressively slow things to a crawl.

Ok, I might be just fueling the fire here, but I feel it needs to be said.

Sure modern windows seems more bloated than before, but not all of these extra features are useless. My favorite features of Vista/7 have been the tiling windows (which can be controlled from the keyboard), the new taskbar (far nicer for manipulating large numbers of windows), and the searching start menu.

Other really big additions are related to security, a must consididering how modern computers are used (The internet is everywhere, sure it sometimes reduces productivity... I'm typing this while I should be working, but it adds far more collaboration). I recently replaced an aging XP machine with a new Win7 box, because the XP machine had been nearly completely owned by spyware that wouldn't have been able to get a foothold on a more recent OS (due to newer security practices introduced post-XP).

So, yes there is generally more bloat introduced with new OSes, but that bloat is usually there to make life easier for the user (more secure, and more intuitive), not just for the sake of using up more resources.

One would think a community of kernel developers would know more about the developments between versions of Windows than changes to the UI.

The kernel hasn't changed much, because it's work isn't too much. Under Windows most of the core components (file system, graphics, UI etc.) are done in userspace. The job of the Windows kernel is to provide architecture specific functions. The kernel is linked against HAL.DLL which does the actual hardware abstraction. That's why we have different kernels for different computers, NTKRNLPA (With PAE), NTKRNLMP with SMP etc. The kernel provides some memory and process management.

I flew off my chair while reading this.
Would you like me to elaborate about the 'etc.' part?

I believe that is untrue. But you might like to view the videos here: http://architects.dzone.com/articles/wi ... ely-faster

Well Performance can also be accelerated by improving the quality of the userspace. I forgot to mention that the speculation that I made about the kernel can be questionable. Seeing that there is acceleration with the boot time too, I believe they might have also been fixing (or optimising) the file system drivers, or the Windows Boot Manager, I don't have proper evidence to prove that though.
And yes the NT Kernel that is for Windows 8 may have gone through a huge change (ARM), I was talking about the changes from Windows 2000 to Windows 7. I should've been clear with that.

And let's blame the user for pushing it so hard. it's 23 April already.

Still not true. The Microsoft developers report some very significant changes in the Windows 7 kernel. For example: http://www.osnews.com/story/22501/Micro ... 7_s_Kernel

I'm sure the same was true for Vista and XP.

Until all agree on definitions of "changed much" and "changed little", it's a pointless debate.

Whilst there is a certain amount of truth in your semantic argument (and, of course, we can't examine the kernel code ourselves) I believe that the Windows developers provide a substantial case to argue that there have been significant improvements in both the Windows 7 and Windows 8 kernels. It would be fairly surprising if there hadn't been such changes since Windows 2000.

But if you think semantic niceties are more important that's fine by me.

I have seen some parts of Windows Vista, 7 and 8 code (kernel and not). Would you take my word? Would the OP take my word? 'cause I can't show the code to you. If you want to see it yourself, get a job at Microsoft as I did in 2005 or see if you can find bootleg source code online.

The Windows kernel was completely rewritten between Windows 3 and Windows 8. There is no sane definition of "hasn't changed much" that is true for this situation. Even between Windows 2000 and 8, there have been redesigns and additions of several systems, sometimes multiple times- graphics drivers, the rest of the graphics stack, sound, user account permissions, assorted security features (ASLR, firewall, etc), solid state drives, networking (especially IPv6), etc.

With all due respect, no I wouldn't take your word as against that of Mark Russinovich. I have proof that he knows what he is talking about; I have no such evidence for you.

Why would you want to take my word against his? Have I somehow implied the kernel has changed little? Knowing first-hand how much needed to be changed and done to make Hyper-V possible? Or do I not understand the use of the word against in the context?

At any rate, the info about improvements in Windows 7 (kernel and user parts) has been available for some time (several years) now, and there are some good things about Windows 8 (not everything is good about it, clearly, but there are good parts, nonetheless). And there obviously were similar things to note and to be said about Windows XP. The OP has likely had many opportunities to learn and reflect about those changes and improvements, but hasn't done so or has somehow failed in the process and has drawn their conclusions. Do you think the OP would now start listening to anyone of reason, with a name or without a name and change their mind?