We have learnt quite a lot about Windows 7 this week, and one of the things was that Windows 7 would not get a new kernel. The call for a new kernel has been made a few times on the internet, but anyone with a bit more insight into Windows’ kernel knows that there is absolutely no need to write a new kernel for Windows – the problems with Windows lie in userland, not kernelland. While the authenticity of the Shipping Seven blog is not undisputed, the blogger makes some very excellent points regarding the kernel matter.I have written numerous times that there is nothing wrong with the Windows NT kernel currently powering about 90% of the world’s desktops. It provides advanced security features, it’s extremely stable, very portable, and supports just about any piece of hardware in the x86 desktop and server markets. “In conclusion, scrapping Windows NT would be a pointless exercise. It is a mature, stable, and, yes, secure system by design.” I wrote a year ago, “Do not make the mistake of thinking that simply because Microsoft refused to enforce proper security policies from the get-go, that NT is an insecure system by design.”
The Shipping Seven blog seems to agree with this position. “The 2000/XP/Vista/Windows 7 kernel has evolved quite nicely over the years, and (in my opinion) is one of the best bits of code we have,” the blogger writes, “You do not, and should not, do a total rewrite of a kernel in a widely-used OS between releases.” He makes a valid point: how often, exactly, has the Linux kernel been rewritten?
It also seems as if a lot of people are confused by the concept of MinWin, introduced by Eric Traut in October last year. MinWin is not a new kernel. It is not a ground-up rewrite of the Windows NT kernel. MinWin is just as he described it: the Windows NT kernel stripped of a lot of things. Nothing new, nothing fancy. Traut also specifically mentioned that MinWin would not become a product in and of itself.
The Shipping Seven blogger reiterates that you already more or less have MinWin; theoretically, you could strip Windows Vista down to just the NT kernel using the command line. Personally, I have no idea how you would go about that, but the authors of vLite do. Sadly, this tool is unsupported by Microsoft.
I agree with the comments on the blog post: Microsoft should provide a tool similar to vLite for Windows Vista – in fact, Microsoft could even buy out the vLite people, seeing as the company has done so before with other tools they found useful. Bless the tool, help improve it, and give us the ability to control every aspect of Windows.
If it’s already so componetised as they say, what’s the big deal?
the problems with Windows lie in userland, not kernelland
Really? Didn’t know the BSOD ran as a user process.
Also didn’t know that process management was a userland thing either. That must explain why when I tell the OS to kill a process, I have to tell it 15 times before it dies. This explains a lot.
Is the whole shutting down process a userland thing too? You know, when you say you want to shut down, then you pack up your laptop only to find that the next time you want to use it you have a dead battery because Windows was telling you “Some crap is not responding…Ok?”
Why are you telling me this…I don’t care if it isn’t responding, I’M SHUTTING DOWN you idiot OS!
The problems with Windows lie in Windows!
Edited 2008-05-30 15:24 UTC
BSOD? What’s that? Never seen one with Vista. In fact, I think I saw one in 5 years of working with dozens of XP systems.
Taskmanager will always kill a process immediately, but will be more likely to result in loss of data. The “End Now” dialog is a bit more conservative, and also invokes the various (slow) crash logging processes. I two wish the “End Now” dialog actually did what it said, but taskmanager is more than adequate.
You shutdown your laptop? I just close the lid, it goes to standby, and then hibernates after few minutes.
As for the “some crap is not responding” – would you prefer the OS terminate applications and lose your data without confirmation? I think I’d rather have depleted battery.
What I cannot understand is why the shutdown process is such a frigging pain – and it has mostly to do with userland programs that pop up interactive dialogs on shutdown. The OS simply can’t know what to do in this instance. It’d be nice if most programs were written to simply save themselves to a safe state on shutdown, and address any outstanding issues when the user next starts the program – “When you last shutdown, you were editing this file but didn’t save it, do you want to save your changes or revert?”
This is as far as I bothered to read your post since you are obviously lying through your teeth. Even if the OS were perfect, hardware issues would have caused more than one if you have really spent “5 years of working with dozens of XP systems”.
Learn to construct your untruths more credibly if you want to be believed.
Edited 2008-05-30 16:05 UTC
Wow. But sorry, I am telling the truth. I had a Windows 2000 system that BSOD’d a couple times on me, I believe due to hardware issues, though I am not sure, but I am not lying about XP – I remember at most one BSOD with XP, though I might be misremembering that one.
More recently I have a work laptop with XP I used for two years – never BSOD’d. I have a work desktop with XP, 3+ years old, never BSOD’d. I have another work laptop I’ve been using for 6 months with XP – never BSOD’d. I have a Vista laptop I’ve been using for 1.5 years, never BSOD’d. I have a Vista desktop I’ve been using for over a year, never BSOD’d.
But then again I never get infected by viruses or malware, a point which credulous cretins such as yourself also refuse to believe, but I think malware and viruses will significantly increase your chances of BSODing.
It seems that Windows bashing is quite popular here and sadly just for the sake of it.
I, too, haven’t had XP BSOD on me, although I take care of about 10-15 systems (those at work, and some of a few friends and family) except a couple of times when dealing with faulty hardware.
To the people that bash Windows: please tell me of an OS as good as Windows, with as many mature applications as Windows (goodbye Linux) and which runs on 300$ hardware (goodbye Mac OS X).
I’m not saying it’s perfect. I’m saying it has its merits but you guys fail to see them out of some reason that only you can understand.
Linux is a kernel, not an OS, so that’s a moot point.
Ooh, way to nail him on that one. Except that everybody connects the term “Linux” with “any operating system that runs on the Linux kernel.” Everybody that doesn’t live in their parent’s basement, anyway.
Edited 2008-05-30 17:47 UTC
but there are a lot of OS’s that run on the linux kernel. 2 people can have 2 very different understandings of linux
Linux is a kernel, not an OS…
Smart ass, aren’t you?
Edited 2008-05-30 18:55 UTC
I’ve had several XP machines here, at work, BSOD on me. But it has always been hardware issues. Such as bad memory or hard drives.
So you complain about those who bash Windows by bashing Linux based OS.. nice.. that’s how to get your point across
I haven’t seen a blue screen of death in many many years. So I’m with everybody else on this one. Windows is a whole lot better than it use to be.
To me not using Microsoft products is more of a political/religious thing. I just simply don’t like the company or the way it conducts business.
Linux, Unix, BSD, and the MacOS X are all equal to Windows IMO. I’d say the Windows platform wins in application support, but that doesn’t, by itself, make it a good OS or platform. Just the most widely used and supported.
Macs are damn expensive, but then again Windows Vista is, imo, way overpriced too. I like Apple, but I truly believe given the opportunity they would be just as “evil” as Microsoft.
People can say what they want, but at the end of the day the majority of people are still going to be using Windows. Linux might be making inroads in the server and desktop market, but it’ll probably be awhile before they “take over”. On the plus side it’s highly unlikely that the Mac platform will ever become the dominate platform which I think suites Steve Jobs.
Im gonna challenge you on that one. I would argue that Linux has near that number of mature apps that Windows does. I would suggest that like many you except that they don’t exist because they don’t share a name with what you are presently used too. Also Windows Vista is not capable of running (properly) on 300 dollar hardware.
For the record this is not Windows bashing. I have no issues with Windows or the people who use it.
“This is as far as I bothered to read your post since you are obviously lying through your teeth. Even if the OS were perfect, hardware issues would have caused more than one if you have really spent “5 years of working with dozens of XP systems”.
Learn to construct your untruths more credibly if you want to be believed.”
Easy champ. I enjoy reading your posts, and agree with a whole lot of them. Does that mean I am lying when I say I have only seen 1 BSOD on Windows XP since the product was released? The machines I maintain, about 200, have never BSOD. The 1 I refer to was when I was experimenting and doing my best to make it blue screen, and was successful. BSOD is not common on Windows XP, though can be done.
So judging byu that last post, you are using TWO accounts on this site ?
is that not against the sites terms and conditions ?
“So judging byu that last post, you are using TWO accounts on this site ?
is that not against the sites terms and conditions ?”
Huh? I have always been DrillSgt for years on this site. Where do you get the idea I have 2 accounts? Any of OSNews staff can look and see I only have 1. What brought this about?
just by reading your response to sbergman27, it looks like you are responding by citing the original poster as yourself.
maybe it is the way you have worded it, but it looks like you are replying as if you were Joshy.
I am probably way wrong indeed, but it just looks like that, have a read over it yourself and you will see what I mean.
“maybe it is the way you have worded it, but it looks like you are replying as if you were Joshy.
I am probably way wrong indeed, but it just looks like that, have a read over it yourself and you will see what I mean.”
Okay, thanks for clarifying. No, I am not Joshy, though I do see how that could possibly be construed. All in the wording I used.
Afraid I am going to have to agree with joshv on the BSOD issue. I use Windows XP, Windows 2000 and 2003 Server at work and XP at home and I cannot remember the last time I saw a BSOD.
Now I have had issues with misbehaving applications (Outlook/Exchange) that would lock up the system, this is further complicated if you use CAC middleware on your machine (DoD). And the result of those problems have required myself and others to perform a “one finger salute”, but I think calling joshv a liar is just wrong. If you have some data to backup that joshv is wrong I’d love to see it.
My God, you’re obnoxious.
In what way? I said that hardware issues alone would likely be responsible for more than one BSOD for 25 machines in over five years even if the OS were perfect. I seriously doubt his claim on that basis, regardless of the name or quality of the OS. I am, however, making some reasonable assumptions about the quality of the hardware based upon the fact that XP is the OS referenced rather than 2008 Server, RHEL, AIX, or Solaris.
It’s how you said it. Look, I know that you’re a smart guy and you express a lot of opinions here. Everybody knows that occasional BSODs happen. Bad drivers do cause problems. But, in my experience, it happens very seldom — and calling the guy a liar because he says he’s been fortunate enough to avoid them isn’t productive or fair. You don’t know what kind of hardware/drivers combination he uses. Your mileage will vary. Not may. WILL. So, lighten up and try to have a reasonable discussion. Nobody gets points for obnoxious behavior. 😉
Are you, Tomcat, presuming to give me a lecture on the value of tact and diplomacy? That’s quite remarkable.
We may disagree sometimes, but I would never presume to call you a liar.
Yes, I agree with Tomcat…. hmmm did I just say that ?
I mean, I agree with Tomcat, that, no matter how heated the discussion, he never makes it personal.
However, I also disagree with the original poster. I support over 600 manchines at work, using a mix of 2k, xp and Vista.
They are a mixture of Dell, HP and Lenovo machines.
They will blue screen occasionally.
We have them all locked down so the users cannot install software, use USB drives, run from CD, they are not connected to the internet, and the applications are MS Office, Notes and Seibel.
They still blue screen.
Standard hardware, standard installs, NO spyware/virus/trojans and still they die.
Edited 2008-05-31 12:18 UTC
Now that you mention it I don’t think I have seen a BSOD in about that long. Maybe the original author is just working on really crappy hardware.
Edited 2008-05-30 18:43 UTC
I dunno, I could second that. I’ve had 3 xp systems in my house since the release of xp, 2 have been perfectly stable with ZERO bsods in 5 years, and 1 has had them weekly. It really comes down to driver support, and if the parent’s dozen computers were all of the same model and that model just happened to have a good combination of hardware, no bsods in 5 years is quite doable IME.
The most stable system for me has been my old gateway notebook, the least stable has been my thinkpad t60.
How do you know he’s lying? What, are you so all-knowing in your arrogance that you presume to read people’s minds too? Give me a break! He didn’t say what he did with the dozens of XP boxes he’s worked with over the past five years so, it is indeed possible he’s never seen a BSOD in that time. Grow up!
BSOD? What’s that? Never seen one with Vista. In fact, I think I saw one in 5 years of working with dozens of XP systems.
I think I have seen my XP installation BSOD once this year. Not much really. But then again, I have had quite a few times when XP just stops responding, or some app stops responding and won’t shut down even from the task manager. In those cases it won’t help even if you try to reboot or shutdown your machine, it’ll just sit there for all eternity. The only solution to such is the power button..
Taskmanager will always kill a process immediately,
As I said above, it doesn’t. I have had a whole load of times some app just refuses to shut down no matter what I do and then it’s also impossible to reboot the system without pressing reset button.
As for the “some crap is not responding” – would you prefer the OS terminate applications and lose your data without confirmation? I think I’d rather have depleted battery.
Well..in the case of battery getting depleted you will lose the application data anyway.
No, any laptop I’ve ever used will hibernate as a last resort when the battery is low.
“BSOD? What’s that? Never seen one with Vista. In fact, I think I saw one in 5 years of working with dozens of XP systems.”
Lucky (or is that “lying”?) bastard. I saw a BSOD just recently in XP, on a brand-new Dell laptop that came *with* it. And aside from that, the number of BSODs I’ve seen in XP over the years on various systems is uncountable; there have just been too many. Vista? Don’t know, I don’t own a computer with Vista, and if I did, it would be wiped immediately for a distro of Linux, and maybe a WinXP dual boot configuration.
“As for the “some crap is not responding” – would you prefer the OS terminate applications and lose your data without confirmation? I think I’d rather have depleted battery.”
Either way, dead battery or the “not responding” message, there’s a chance you’ll get data loss. Also, in my own experience, I’ve noticed that these “non-responding” programs at shutdown time have absolutely nothing to do with what I was doing, and just killing them would have caused no data loss in any cases that I can recall. Besides… you would think that when telling a computer to shut down, you would have made sure that anything important has been saved to disk… right?
So you’ve never used Vista, and never will, but you hate it?
Well that’s simply not true. If you shutdown and apps or services are not responding, Windows shows the dimmed screen with the list of processes preventing it from shutting down. If you don’t answer, it will eventually just shutdown. Then again, I haven’t actually seen this screen since the beta, so I’m rather skeptical about it being a common occurrence.
Besides, even if it’s stuck and you close the laptop, it will go to sleep. XP may have had some problems with that (mainly due to crapping add-on software from IHVs for docking and such), but Vista doesn’t.
I have used Vista for two hours on my new laptop. Finally I wiped the harddisk and installed Ubuntu.
Though i have used microsoft products since 3.11 for workgroups. Perhaps Vista is a different breed but XP included, stopping an application or a hung process isn’t as direct and straight forward as it is on on Unix.
It would be nice if from the cli you could “ps -waux | grep <process/app>” and kill -9.
On linux,freebsd,.. i never witnessed a hung cdrom choking the whole system.
taskkill /f /im <name of your app>
The only way an executable can stay running after you terminate it like this is if it is stuck processing I/Os in a driver. If that’s happening and you really want to learn why, obtain a copy of windbg, start a livekd session, run “.symfix” and “.reload” to get symbols, find your target process and thread using “!process 0 7 <name of process>” and if you see any IRPs mentioned in the thread description, run “!irp <address of irp>” on them to find out which driver is holding you hostage.
But should the explorer hang as well due to that particular misbehaving driver?
Is the driver is misbehaving, the kernel cannot do much about it most of the time (in so called monolithic kernels, at least). If the misbehaving driver for example corrupts any data structure used by the kernel, you are screwed.
The problem really is misbehaving drivers, and the fact that a lot of things used to run in kernel space in windows did not help, as well as the big number of crappy drivers. Fortunately, with Vista, it looks like they have put a lot of things in userland, where it belongs (that’s what they pretend, at least, I have not checked it, I am not interested in windows). Less code in kernel space certainly means more reliability here.
1) I don’t think most explorer hangs are actually driver-related, except maybe in an indirect way. The old explorer used to do actions which were cheap most of the time but could occasionally be really expensive on the UI thread. That’s of course quite bad. Vista is a lot better in this regard and I have yet to observe such a file browser hang (it also helps that I never use CD-ROMS anymore, but I do use network shares exensively).
2) Vista moved a limited set of drivers out of kernel mode. Mostly it was the ones that seem to cause the largest number of crash hits: Video, Audio, and some random USB devices. Moving stuff up into usermode has a ost too, though: complexity and poorer performance due to more context switches. This was able to be avoided for video through some pretty cool changes to memory management and an interesting buffering scheme. But the complexity of moving things up to usermode sometimes means that your device is less likely to bring down the system but also somewhat more likely to not function correctly in low-resources or some other scenario in which the user-application fails to communicate properly with the user mode driver host.
Like with everything, there’s a tradeoff. If your code is in the kernel and it’s buggy, the box will blue screen and you have no choice but the fix those bugs post-haste. There’s a lot more tolerance for lower-quality user mode code, so one could imagine drivers becoming buggier since the individual bugs are of less consequence.
This is good information to have, going to have to save this to a notes file somewhere.
I wonder if Process Explorer is doing something similar when it is able to kill tasks that Task Manager can’t.
This is probably because Vista and XP are set to auto reboot when a system error happens.
One thing that is sure. XP and Vista don’t have BSOD as much as Windows 9x. But BSOD still happen on occasion.
I haven’t seen XP drop a BSOD not identifiably related to bum hardware in almost 5 years. And, though it pains me to admit it, I’ve never seen Vista BSOD… ever. I have not seen Windows 2008 BSOD yet despite many connected users over terminal services.
The Windows kernel has been significantly hardened in the last few years, and it shows. Those who complain otherwise about issues that haven’t been issues in the better part of a decade sound like people who complain motivated on politics with absolutely no experience with the technologies they complain about.
Early on with Vista I had what probably would have been a BSOD in XP. I installed the latest and greatest Vista nVidia driver (my mistake) and the driver itself would quite frequently crash while playing games. In XP that would have been a BSOD – in Vista the game crashed, and the driver restarted.
These days though I haven’t had a video driver crash in probably 6 or more months. nVidia has cleaned up their drivers. No doubt the fact that the system doesn’t blue screen when the vid driver goes down helps them troubleshoot the issue.
Very true, Vista does handle more than just video driver crashes much better as well, My sound card drivers used to crash constantly (creative) when vista was just released, but it never BSOD’ed on me, and I could even get sound back by disabling and re-enabling the device, not so on XP, and not so on Linux, which would panic or just freeze all together. While windows has made alot of mistakes, it is becoming quite stable (havent shut down my vista media computer in many many months). Many companies release hardware products that were brilliantly engineered, and hired the cheapest programming team to write the drivers, and its nice to have a kernel robust enough to handle thier mistakes so you can keep on going on.
Exactly. It *is* possible to BSOD Vista, but this is entirely due to driver faults, not the O.S. kernel itself.
Case-in-point: download the latest Razer drivers (if you have a Razer mouse). Terminate one of the smaller child processes like razerhid.exe or razerofa.exe through Task Manager. Then open up the main driver control panel and change a few settings. You won’t get past “Apply” or “OK” without a BSOD, guaranteed.
Should I blame Microsoft for this error? Hell no! The blame lies solely with the peripheral maker and its buggy driver software. Thankfully the Lachesis mouse works fine with a generic driver, which is what I use from now on.
Edited 2008-05-30 22:41 UTC
Lol, how can someone say such nonsense and try to sound serious.
Do you even know what a computer is?
No it’s not the thing that looks like a tv (that’s your monitor).
A good kernel doesn’t let userland apps completely paralyze the whole system.
When it comes to Windows, even a freaking Notepad can stall a system (just try to open a really big text file, see how responsive your system becomes).
The only thing hardened about Windows kernel is the users fist when they sit helplessly in front of their computer waiting for the hourglass to turn into a pointer again.
The first Vista machine I used BSODed continuously, mind you it was when the thing had just been released and was running in a crazy laptop. XP has never BSODed on me but it has restarted a few times because of faulty hardware, yet nothing that can be tracked to Microsoft.
You can be mad at them because of their closed-sourcedness, their mafia-pushed standards, the stupid kids they have designing their GUIs, DRM, or the compulsory activation of software as I am, but you(the original flame post) lose all your credibility when you start outright lying about their software and its faults, denying problems with Linux, and other similar FUD tactics. You have become them, YOU ARE MICROSOFT.
The NT kernel is a modern kernel, it is similar to the Linux kernel in some ways, it offers similar features and has been updated as well to include the latest trends in lower end computing. A faulty driver or hardware will screw them both, BSOD or otherwise. FUD from the Windows 95 era is not fun or accurate anymore, so, stop it.
And I thought this site had a better population than Slashdot. In the end the GNUyatolahs are there to spoil any intelligent discussion.
We like OSs and want to discuss the real technical issues and differences between them and to get excited at the prospect of new better systems even if at the end we are gonna be disappointed.
The only time I had Windows XP BSOD was with a bad driver with the pinnacle pctv card I bought – oh, and in Windows NT4 because of a memory module I had which was faulty (and caused issues in Linux as well as Windows 9x too).
The problem is that far too many people here have some sort of emotional baggage that comes with a given operating system – as if the operating system were an extension of their personality, and by virtue of critiquing their operating system, they see it as an attack on them.
Well, don’t expect a decent level of conversation here; it is as bad as Digg, Slashdot, Kryoshin and other troll related boards. Arstechnica used to be interesting the battle front until you have the uneducated rantings of cluelessness from the cheap seats.
I think I can say you speak for all of us, victims of Microsoft.
Now please ignore the posts form the Microsoft paid shills that will say that they never get any BSOD, and their Windows systems have 1000 apps installed, and running 100 apps at the same time all multitasking smoothly, and have uptime of 5 years without any slow down.
I’m a Solaris administrator, not a paid Microsoft shill. based on my 10 years of experience in working with and around Windows systems, I can count the number of BSOD’s I have seen on two hands. If you are going to bash Windows, at least come up with something that actually happens on a regular basis that would constitute a problem for a large number of users.
I am sure my experiences are not unique, and as others have pointed out poor drivers have a great deal to do with various Windows issues. If you have nothing of substance to add to this discussion other than your anti-Microsoft nonsense, either stop trolling or go somewhere else.
Hilarious. So, people that aren’t experiencing BSODs are “Microsoft paid shills” now. Microsoft must have millions of people on its payroll, then… LMAO!
Edited 2008-05-30 20:34 UTC
I’ve been saying that I have no BSODs, and I don’t. I am also not some sort of paid shill. I’ve had my share of troubles with Vista. I have been hit by some odd search indexer and start menu indexer bugs that caused the search engine to thrash the disk and consume 100% of CPU – only on my desktop though, not my Vista laptop, and it ran fine for a year before the issue.
I disabled search indexing and set my start menu search options to “Don’t Search for Files”. Which really is a pretty unacceptable solution. Search indexing is a major feature of the OS, I shouldn’t have to disable it entirely for some inexplicable reason.
So I have had problems. But this most definitely isn’t an issue with the kernel.
Your comments are so wrong about process scheduler. When you press the close button on the window of a process, it sends some windows message to that process. It then waits for some time (sorry i don’t remember exact time) and if the application doesn’t respond then the dialog box is shown to the user to terminate and send report to Microosft or just terminate and not send report. This is the slow process.
Instead, just open task manager and kill the offending process and it will die in less than 1 second.
He claimed that he regularly has processes that are not killable even in the task manager. I *think* I’ve run into this behavior once or twice – and procexp from Sysinternals was able to kill the process, but otherwise, 99.9% of the time task manager will do the job.
Going to have to say that I have ran into many processes that task manager just would not kill for one reason or another, but Process Explorer had no issues. Sometimes task manager kills immediately, sometimes after about 1-10 minutes you get the “End Now” dialog, and sometimes that process just happily stays running in the background after telling the task manager where to stick it. I will grant that this latter situation is in the minority, but it definitely does happen. Before discovering Process Explorer the only way to get rid of these rogue pids was a reboot.
I want to say that the kill.exe command would get rid of them as well, but I do not recall if it came standard or if it was something I installed from a resource kit or powertools. I would go look but I don’t have any Windows installations left. I have two fully legitimate licenses for Windows XP (not counting any prior versions), and no installed copies. Won’t be installing Vista on any machine I own anytime soon.
windows comes by default with the ‘taskkill’ cli command.
except when it doesent die in less than 1 second, and it takes minutes….
Why in the world would a user need or want to strip his/her os down to command line only? That is something you would want to do in the server room, and you can with 2k8.
Something to keep in mind with MS products is that they are a big fan of the specialized sku. When it comes to install process, normal users want a LESS complected one, not more complected. While the internet geeks cry for more uninstall options, how successful has Windows-N been in EU? Normal people don’t care about these things.
What MS should be doing is adding a geek sku to the lineup. 2k8 is incredible, but a) you don’t want to pay 3k for an os, and b) you don’t need half the things that make it cost so much (app server, domain controller, etc).
One of the comments in the linked blog entry I think gets to the heart of it. Nobody really wants to do it for practical use, but it’s the capability that makes geeks wet their pants.
If given a “minwin” install option, most geeks will try it out, boot to a command line prompt, launch “sol.exe” and go “cool!” and then write a blog post that MS has finally gotten out all the bloat. Then they will reinstall with all the crap actually required to run real programs.
More practically somebody suggested a “remove components required for backward compatibility” option. That’s an awfully nebulous concept, but even if MS could do it, they wouldn’t, because there would be no noticeable performance increase. I suggest MS try – so that the geeks can do it, report imaginary performance increases in their blogs (even though half their programs don’t work), and then reinstall the backwards compatibility components so that their programs work again.
Luckily MS doesn’t listen to geek blogs when deciding product features. Most people could give two craps about slimming down Vista – all they want it to do is run their programs.
actually, the last paragraph of my comment is
Vista is the only windows so far I haven’t hated, but 2k8 is the first version where it starts entering my list of favorite operating systems.
Edited 2008-05-30 15:58 UTC
”
While the internet geeks cry for more uninstall options, how successful has Windows-N been in EU?
”
Would that be the Windows version where they removed features but then still sold it at the same price on the shelf along side the regular version then claimed N was a flop because everyone chose the other box? Would that be the Windows N failure in the EU market you refer too?
What i mean is that people want what they are already getting, they want as little control panel configuration as possible, let alone having to go out on the web and download something new.
The people who complain about the lack of vista modularity are not the target audience of the sku. It has nothing to do with the ability to ship it, because it exists in 2k8. The point of my post is that they should ship a stripped down 2k8 for the geeks and charge the same price as the home version.
I wouldn’t mind if MS had a “Developer SKU” or something like that, which had as little cruft as possible. Probably distributed solely through its MSDN subscription service. That would make the geeks happy, without confusing customers.
Would that be the Windows version where they removed features but then still sold it at the same price on the shelf along side the regular version then claimed N was a flop because everyone chose the other box? Would that be the Windows N failure in the EU market you refer too?
Did you forget it was EU that ordered MS to set the same price for both versions?
Anyway, if I wanted Windows without Media Player, I would not care that other Windows, with Media Player, is priced the same. Why would I? I’d simply buy the one without it because I wanted/needed Windows with no media player.
So yes, Windows N is something no end customer really wants.. Microsoft’s competitors are different story. Don’t confuse those two groups.
Edited 2008-05-30 21:33 UTC
Was it the EU that required MS sell a crippled Windows (by comparison to there existing line) for the same price? Ha.. If so, I’d missed that bit of information.
Maybe MS should have retained tehre media player, included uninstall and provided optional installs of competing media players.
“The call for a new kernel has been made a few times on the internet, but anyone with a bit more insight into Windows’ kernel knows that there is absolutely no need to write a new kernel for Windows – the problems with Windows lie in userland, not kernelland.”
Is this still as true as it used to be in the days of Vista? In the days of an ultra-paranoid kernel wasting half its time making sure a user doesn’t try viewing or listening to the contents of a media file just in case it’s “not authorized”? I recall hearing this stuff was so deeply entrenched in the kernel, the chances of Microsoft even bothering to remove this was next to zero (although I suspect the big reason is they already promised the big **AAs).
I honestly do not believe ANYTHING about DRM has any business deep down in the kernel… but that’s the way Windows is going. Good thing there’s Linux. And yes, I would take the inability to play that crap, period, over having my computer actively trying to police me. Never did give a damn about movies for the most part, and thankfully, audio CDs aren’t mangled with DRM.
I work on the Windows Kernel and I can tell you that no piece of NT that actively snoops on what you’re doing and tries to protect content-proucers’ rights. There are a few features (I can think of 3 of them), which make implementing a DRM system in user programs and drivers a little easier, but none of them have a performance impact when not in use.
Here’s one that actually helps performance: Protected Processes are created as a unit from kernel mode, unlike the normal processes in older versions of Windows, where they are created piecemeal through cross-address-space writes. The new method is actually faster, so all process creation goes through the new path (just without the protected process flag).
//sarcasm;
Phooey on your facts. Winblows! MicroShaft! Big Brother! etc. etc. etc.
//sarcasm;
See, I was right.
There are Microsoft paid shills working here.
And it’s funny how poetically they are able to explain the crapiness of Windows kernel.
Did you go to PR school or computer science school?
I guess when it comes to Microsoft either one qualifies you to work on Windows kernel.
It’s easy to make Window sound bad when you make things up.
Have you ever, ever heard of ANYONE being unable to play a non-DRM’d media file on a Windows PC? Of course not. The only DRM lives in the *PLAYER* – and it only applies when you buy DRM’d content. If you don’t want DRM, don’t buy any DRM’d content. Simple as that.
Of course the kernel hasn’t got anything to do with DRM. How about you think for a second before posting your mindless Linux propaganda?
Edited 2008-05-31 17:10 UTC
My take aways:
1. Modern Windows consist of a functionally mininal “winmin” kernel that is some of the best code MS has.
2. To produce an OS for a particular purpose, the kernel is simply adorned with a suitable selection of feature components that can be attached via XML.
3. Windows is actually a highly modular and “componentized” OS.
From this, I conclude from experience with Vista:
1. The choice of components that make up Vista was very poor in many respects. The components add or detract from previous functionality, are poorly coded and/or designed and perform very poorly.
I therefore recommend the following to MS:
Rethink the component selection that goes into Windows 7 for the categorical uses, desktop, server, mobil, etc.
Allow the system to respond to users needs for new functionality. That is, default installs provide 90% of the users with 90% of the functionality for that particular use/platform. Let the so-called 10%’ers to leverage the cited XML routine so that the system can dynamically add/enable needed components and features.
Following the above logic, do away with the various flavors of windows (i.e, standard, premium, ultimate, et. al.) and just give us one, dynamically adaptable Windows that serves MY needs. Call this new radical OS, Windows.
Finally, clean up the design and performance of the components we use most. The ones that cause Vista to get such a bad rap. You are already doing this, I can see, as gaming performance now approximates Windows circa 2001. That’s a good step but we want better performance, not to be simply releived when we par on decade old performance metrics.
Yeah right. And lose one of the most central ways to gouge their customers for more $? Come on. Windows servers, in their standard editions, can barely take advantage of most modern hardware until you drop another $1500-$3000 dollars.
Hmm so far so good?
Nowhere near NTX i guess.
Microsoft’s own feedback about 16% were OS crashes with Vista, 30% nvidia. Bad drivers have always been the issue with Windows but one should look beyond that, how the OS handles that driver crash.
As I remember Vista was supposed to handle driver crashes better by just restarting the driver, from what I’ve seen it makes the system completely useless and a restart is required anyway.
As I mentioned in another post, I’ve had crappy nVidia drivers crash quite frequently while playing games. The driver restarted successfully, though the game was dead. Never had to reboot as a result. These days though my nVidia drivers don’t crash any longer.
Gates et al have always claimed dominion over the “Windows Experience”. At first they were able to get away with it, but now even nontechnical people, as witnessed by the comments to the blogger, are starting to see that they should have some say in how much of and what parts of an OS is installed on the machines that they own. I’m sure that Microsoft does not want to open up the can of worms an installer similar to a Linux install would cause, but in the Linux world you have the choice to install everything or pick and choose which packages you want. Does this cause problems? Of course, but it’s my machine and if I screw it up, I get to fix it. PC’s have been around a while now and similar to the maturation process in general, just saying “Because I said so.” only works up to a point. Users are growing up; Microsoft needs to mature or lose more market share.
Disclaimer: I use Vista at home to play games and XP at work for email. Everything else I do on Linux. I think that pretty much sums up the capabilities of the two environments.
Later.
It should actually be WaxWin… the kernel that melts when you use it
No really there is nothing wrong with NT, just with who and how they use it.
I agree with Thom. The kernel for WIndows NT, based on Dave Cutler’s work, is sufficient to do the job. The problem lies with backwards compatibility, and the userland design in reference to compatibility.
I understand some of the comments made about businesses needing legacy programs; however, software isn’t like a wrench that you can use for 15-20 years before it breaks. It’s not that kind of tool.
In my estimation, Microsoft best course of action would be to sell 4 versions of Windows; Windows 7 Home, Windows 7 Business, Windows Legacy Home, and Windows Legacy Business.
The two legacy products will be able to run all software from 3.1 all the way to XP SP3. Legacy would be coded to prevent Vista software from loading for the majority (not that hacking wouldn’t occur).
Windows 7 Home and Windows 7 Business would be legacy free code, which could be optimized to run based on clean, efficient code. Provide an SDK with HIG guidelines like Apple does with it’s developers, and you might see better written code.
Also, there should be an option for running WIndows Legacy within a VM on 7 Home or 7 Business, so that if you needed to install a product, you could do so, for an additional charge of course.
Otherwise we’re going to run into a 5 year development cycle again when Microsoft finds out they can only extend Vista’s code so far before scrapping it, like they did with XP.
And why not?
If the software didn’t hardware purchases necessary, I doubt many people I know would purchase computers less than every 10 years.
“but anyone with a bit more insight into Windows’ kernel knows that there is absolutely no need to write a new kernel for Windows”
A LOT of people would disagree with that. Windows has tons of problems that are being held up like a house of cards.
The Registry has GOT to go. Not cleaned up or fixed or revamped but completely gotten rid of. Apps need to go back to being self contained in their own individual folder structures accessing only system calls outside of there.
That’s a place for them to start. If they want more they can put a billion dollars in my account and I’ll get back to them eventually…or not.
*sigh*.
And the registry has what to do with the kernel, exactly…?
Please, these issues are confusing enough as it is for most people. Don’t try to make it worse.
The registry is an important subsystem of the kernel, but it is significantly harder to corrupt and more suited for its purpose than using flat files.
The equal statement to this fellow’s assertion is “the /etc/ tree has to go!” I don’t think either of those statements is true.
I don’t see how you can say it is harder to corrupt: if the registry is corrupted, you simply cannot use windows anymore, you have to reinstall it. Difficult to have easier to corrupt: that’s a single point of failure. Which is certainly not the case of /etc. The goal of the registry was to have a single place to handle configuration, but it was because windows did not have a sane way to store configuration (one directory for the system, and $HOME/.foo directories for users).
No, it is not: the problem is not using configuration in a centralized manner, the problem is the implementation of the registry, which is horrible (binary, keeps growing, etc…). But this is arguably a complex topic: /etc is certainly limited once you want to manage many machines in a large network, and it is a shame that every software uses its own configuration format (with different behaivour related to escaping, etc…). The .ini is actually often enough for simple applications. To solve the network adminisation problem, Solaris and mac os X for example have their own way to deal with it in a centralized manner. But /etc is still there.
Anyway, this has not much to do with the kernel.
It has to do with the kernel because the registry implementation is in the kernel, and it is used by the kernel quite early in boot.
The registry is essentially a specialized filesystem. So it’s a single point of failure inasmuch as a filesystem is a single point of failure. The binary format of the registry is pretty much analogous to the format of a filesystem as well (it’s a journalled, transactional data store with per-key ACLs, symlinks, etc). It’s actually implemented in multiple files, and the per-user portions of the registry are stored in separate user hives in the home directories (I think the specific file is ~\ntuser.dat). The system-wide hives needed for booting and managing drivers have backups taken when you make config changes (the “Last Known Good Configuration”), so if the current hive file becomes corrupt due to disk problems, there’s a chance of recovery through the backup files in %windir%\system32\config\regback. I’d say the registry implementation is not so horrible.
Have you had any specific problems with the NT registry? I have no idea what the inside of the 9X registry was like, so perhaps your complaints were justified in the past.
But that’s exactly what’s bad with the implementation: (a fs) means that the configuration became a single point of failure. Also, having to reimplement all this as a fs is kind of stupid: why not using the FS for this ? This gives you more fine grained control (ACL are per key, right ?), but I fail to see how this is useful. This has never been useful on any other OS (and
The fact that it is not documented (applications put way too many things in there and cannot work when it is not there anymore; arguably not the fault of MS here, at least not directly), keeps growing is enough for me to say the implementation is quite horrible.
All the time when I have to use windows to build software for windows. In particular, uninstalling an application always leave garbage in the registry, which is annoying when you install a new version of the same software (visual studio comes to mind: install visual studio 9 and 8 has still keys there, which confuses many softwares).
And it also means it does not work at all to reinstall the OS without reinstalling all the applications, even if they are still there: all the path informations about the software are again in the registry, which means nothing is self contained.
This won’t happen.
If an application is self-contained in its own directory, then there is no need to “install” it as such.
One could just sit at one machine with the application installed, and copy the application’s folder to a USB memory stick. Remove the USB memory stick … drive home … sit at another computer … copy the folder back on to a different hard drive, and away you go.
http://portableapps.com/
The registry is there because it has a design purpose … within a paradigm where users have no rights to copy software for their own use that is.
Edited 2008-05-31 07:15 UTC
Automatically adding/removing services/processes (modules) as needed. Nothing unnecessary running in the background, ever. Nothing unnecessary on the HD.
Scenario 1
You install Windows, everything gets installed. But it’s running nothing. Absolutely nothing. Boot time is seconds. Then you ask it to do something. It starts the process and the needed services. And so it goes. After a while, Windows notices that you don’t ever use certain features and it off-lines them; deletes them from the HD. You can get them back via DVD (or copy on you HD) or the net.
Scenario 2
You install Windows, nothing but the basics gets installed. The DVD is copied to the HD. Boot time is seconds. It’s running nothing. Then you ask it to do something. It installs the needed parts, starts the process and needed services. And so it goes.
Scenario 3
You install Windows, everything gets installed. And it’s running everything. Absolutely everything. Boot time is perhaps minutes. Then as you use it, days, weeks, months, it starts stopping and disabling processes and services you never use. It gets leaner and leaner, faster and faster. And so it goes. Then it starts off-lining them; deletes them from the HD. You can get them back via DVD (or copy on you HD) or the net.
Today, I’m killing every service I don’t need after Windows boots. If I don’t ever need them, I disable them so they never start.
Senario 2 would probably be the most compatible with users, but they are good ideas that all OSes should implement (Computer: I see the user hasn’t used Evolution in the year the system has been running, time to delete it)
The thing I can’t stand the most about many people on here is their constant chanting of ‘MinWin’ as if it were the panacea to all of life’s computing problems. The simple fact of the matter is that it isn’t the solution to all of life’s problems.
WinMin is not a product but an approach to modularising and defining what is the ‘core’ of the operating system – that is, the bare minimum parts of Windows to have so that one can create a customised solution sitting on top. So ‘MinWin’ is not just the kernel. To say it is ‘just the kernel’ is complete ignorance of the whole point of ‘WinMin’.
The side effect of that move was the uncovering of a number of unsavoury programming practices where used in the ‘core’ of the operating system – which is what slowed down to a certain degree some of the Windows development (hence the re-writing of some of the parts of Windows too); there was spaghetti code where parts lower down were pulling in resources higher up the chain. It was a nightmare.
Getting back to the topic in hand; the issues with Windows, however, have nothing to do with the kernel; heck, the registry itself (which someone raised – given it has nothing to do with the kernel, I question its relevancy) in theory doesn’t have anything wrong with it – the problem is that it was abused to such an extent by developers that it got the reputation it has now. There is a good video on Channel9 over the registry and how the issues can be addressed without needing to throw the whole thing out the window and start again.
Now, why don’t they just release a tool that uses those XML files for enterprise users so we can make builds that contain only the software and drivers we need?
> “anyone with a bit more insight into Windows’ kernel knows that there is absolutely no need to write a new kernel for Windows – the problems with Windows lie in userland”
The problem with this statement is that it’s completely untrue. The failure of Microsoft to modernize the Windows NT kernel and instead clinching to legacy code is what is keeping MS ten years behind (Also, the only people with any insight into the Windows NT kernel are MS NT devs and rootkit developer experts).
> “In conclusion, scrapping Windows NT would be a pointless exercise. It is a mature, stable, and, yes, secure system by design.”
Also false. First of all scrapping Windows NT would be the best thing Microsoft could do right now for their Windows product line. And I wouldn’t call the long history of exploitable bugs in Windows NT making it stable, in fact it negates whatever efforts they had made to try make it secure (by design).
> “You do not, and should not, do a total rewrite of a kernel in a widely-used OS between releases.”
And why not? I thought Microsoft had the resources? I guess the complexities of modern OS development have hit reality, and even Microsoft is now realizing they can’t just buy a system and resell it as theirs (think Q-Dos and the 80’s).
> “MinWin is not a new kernel.”
Having actually looked into I can indeed say that it is.
> “Nothing new, nothing fancy.”
Again, untrue. “MinWin” and Singularity is the newest technology MS has right now in the pipeline, and having released it as “Opensource” they have apparently shot themselves in the foot, as they will have a hard time using it under any other name.
> “Microsoft could even buy out the vLite people”
Wouldn’t surprise anyone, this is business as usual for MS.
yes
Where is that source so we can look into it? I have no reason to believe that Microsoft would develop a hot new kernel and not be chanting its goodness and its ability to bring the dead back to life and to make wine from water from years before starting to code a single line of it.
Singularity isn’t anything more than a concept right now. It’s not going to be Win7 nor Win8 nor Win23 for that matter. Maybe some of its ideas make it to Windows proper, but probably none of its code will be reused just because people have seen it. This is how Microsoft works.
I really cannot understand how you fail to see that NT and Linux, and kBSD are all the same thing. They might work differently under the hood, but they expose the same general functionality. A new kernel would be exactly the same. Both Windows and Linux include things in kernel mode that shouldn’t be there in theory, but they are gradually getting them to userspace where they belong.
All a modern kernel must do is to provide a memory manager, a scheduler and a HAL(and now virtualization) for SMP and must be pretty good at it just to boot up. It is very difficult to provide this functionality, but once it’s there in machine independent code you need not worry anymore. Well, you do, but just don’t need to rewrite anything from scratch because you have a working base to improve.
bsod on nt is in 99% driver /hw issue, but u managed to kill my dev win2k machine by creating invalid ip packets. however thid is no excuse either do the drivers yourself or bloody test/audit them etc. for all control freaks use process explorer to kill anythig without having to wait….
“Paid shills”
“PR school or CS == both the same to satisfy prerequisites required to work on the Windows kernel”
What are you trying to prove ?
Are you trying to confirm the notion that many fanboys on the other side of the fence have about *nix/OSS supporters being sort of religious zealots?
Not recognizing the things done RIGHT by MS with Vista is not a way to help open source software (I’d even say it hurts OSS IMHO), it does not STICK IT TO THE MAN.
It is wrong not to recognize mistakes made by closed source or open source projects alike just like it is wrong not to accept the great under-the-hood work done that might or might not fully show to the end user in a flashy way.
I am not going to throw away KDE4 because versions 4.0.0 through 4.0.4 still do not feel polished or close to being complete: I do recognize the great effort spent on all the new core frameworks KDE4 is built upon (Plasma [and the new Kwin], Solid, Phonon, wide QT4.3/4.4 deployement,Akonadi, etc…).
By the same toke, one cannot forget Aero (Flip 3D is a poor replacement of Expose’ and more tricks should be stolen from OSX’s Aqua ? Yes, but Aero can be extended to do just that… we already jumped away from GDI+ and old Windows window management and compositing), the massive sound/video driver model changes (WDDM and the reduction in BSOD’s), WPF, WCF, deep Windows Defender integration, randomized address space layout, Data Execution Protection, IE7’s sandbox mode, UAC (yes, UAC … Mandatory Access Control was sorely lacking as default option in Windows before Vista IMHO), better two way built-in Firewall support, etc…
Yes, some of those features were introduced with XP SP2 and/or have been backported to XP later on, but when you go down to it they are not as deeply integrated with the rest of the OS as they are in Vista (Windows Defender and IE7 Protected mode are two examples of what I am talking about that quickly come to mind).
I dual booted Windows Vista and Windows XP on the same Core Duo+GeForce Go 7300 (64 MB VRAM)+1 GB of DDR2 system (with Fedora installed on the side) and even on that machine after getting accustomed to Aero, Windows XP’s GUI felt a bit too unresponsive to me.
I stand by what I always say to people wondering about downgrading: unless your system does not run Vista reliably or smoothly then leave it on (you are better off with it).
You folks should stop quoting that “Shipping Seven” blog. The stuff he posts is so vague, that it could be anyone, including a crazy Windows fanboy. Until his authority is confirmed, it should be ignored.
I’m going to Rant for my first post on OS News…
Windows is ridiculously bloated.
There’s tons of stuff turned on by default, and tons of stuff bundled, that most people don’t NEED.
~I don’t need 15000 handles open at an Idle Desktop.
~I don’t need 15 GB OS garbage for an Idle Desktop.
~I don’t need my CPU spiking 5%-10% utilization at Idle Desktop.
~I don’t need EVERYTHING Pre-fetched.
~I don’t need 40+ processes running in the background.
~I don’t need tons of MS crap the phones home (Not just talking WGA).
What I do Need is:
~An OS that uses very little resources and dedicates most of my resources to the APPLICATIONS that I run.
~An OS that can use the full 4GB of Physical RAM if 32-Bit, and can dedicate more than 2GB to a single process.
~An OS with no ARTIFICIAL limitations on how many processors I can have or how much RAM I can have.
~An OS that basically ONLY: launches applications, manages memory, multi-tasks well, has network connectivity, manages modern hardware, A basic functional Desktop, and has compatibility with all Windows 32-Bit Applications. PERIOD.
Unfortunately, in the Microsoft World that will never happen. It’s about control. All the Phone-Home-Ware is to have control over your machine. All the Bundled CRAP (IE, Media Player, Messenger, etc.) is to control content and attempt to create vendor-lock.
Other than gaming, I have absolutely NO use for an MS OS, at home (I do everything else on a Mac and Linux Boxes). So what MS needs to make is a stripped down “Gaming” OS (XBox OS + Desktop?) that runs all windows 32-bit Apps. Hell, I’d even pay the “Ultimate Version” price for that.
I’m really getting sick and tired of the “What Intel giveth, MS takes away” paradigm. It’s just stupidity at this point, not “Innovation”. It’s just sloppy-@ss coding that programmers have let become the “Status Quo”, unlike the Mainframe days where quality and tight code was essential… and IT people were actually respected and awed. Nowadays we’re looked upon as “Those overpaid dumb-@ss IT Jerks”, thanks mostly to MS.
1) Do you even know what a handle is? I thought not. That request makes no sense. You don’t even specify what kind of handle you’re talking about. Nevermind the fact that being “idle” has nothing to do with how many window handles / GDI handles / whatever other handles are being held.
2) Are you some how conflating the size of an OS plus libraries with whether the desktop is idle or not? Again this makes no sense. And Windows doesn’t take up 15GB of disk space, either. It takes around 9 or less. A little less than OS X 10.5.
3) If you don’t want your CPU to do anything, turn your computer off. As long as it is running, it is doing things. CPU usage should be as high as possible while some task is being performed, otherwise your CPU is being wasted.
4) That depends. You don’t need 40 processes, but which ones are you going to turn off? Oh wait, you actually do need those. Damn. And more processes is better (better isolation of tasks, improves reliability and security).
5) Why wouldn’t you want things to be cached in your cache (RAM is a cache)? If you don’t want things to be prefetched, just remove some RAM from your PC. Problem solved.
But you just asked not to prefetch things. This is a very confusing request. The OS uses more memory for applications, and you say that’s bad. Then you say what you want instead is for the OS to do what it’s already doing?
This is a perfect example of you not knowing what you’re asking for.
1) You cannot have a 32-bit OS that can use all 4GB of your physical memory. The two are mutually exclusive. 32-bit means 4GB of addresses. Unless the only thing you are addressing is physical memory, you will never be able to address 4GB of it. Since virtually all computers have other things to map addresses to (like, say, a video card) – it’s quite obvious that this is never going to happen.
That’s why we have 64-bit OSes now. If you have 4GB of memory, you should get one of those.
“2) Are you some how conflating the size of an OS plus libraries with whether the desktop is idle or not? Again this makes no sense. And Windows doesn’t take up 15GB of disk space, either. It takes around 9 or less. A little less than OS X 10.5.”
OSX 10.5 has more applications than MS paint or calculator.
“1) Do you even know what a handle is? I thought not. That request makes no sense. You don’t even specify what kind of handle you’re talking about. Nevermind the fact that being “idle” has nothing to do with how many window handles / GDI handles / whatever other handles are being held.”
Handles are basically “connections” to the Windows API. XP runs with 3000 open, Vista runs with 15000 open. It’s a waste of resources, sloppy, and is one of the primary reasons for Vista’s larger CPU spikes at desktop idle.
“2) Are you some how conflating the size of an OS plus libraries with whether the desktop is idle or not? Again this makes no sense. And Windows doesn’t take up 15GB of disk space, either. It takes around 9 or less. A little less than OS X 10.5.”
Exploding the “library” with more spaghetti code is not what I consider efficient. Again, XP runs in under 2 Gigs of install. Vista Takes 4x more + (Depending on the version). Vista doesn’t offer anything special warranting that kind of bloat… including the new “security features”, and the ridiculous 3D UI. At least with OSX the eye candy has purpose and functionality… it doesn’t just sit there looking pretty.
“3) If you don’t want your CPU to do anything, turn your computer off. As long as it is running, it is doing things. CPU usage should be as high as possible while some task is being performed, otherwise your CPU is being wasted.”
That’s the dumbest statement ever. Again, comparing to XP the CPU spikes are infrequent and at about 2%, at desktop idle. Vista has so much “crap” going on in the background, that it takes way more resources than an Idle desktop should. If I haven’t launched anything, it should basically be doing NOTHING (Except some minor housekeeping / memory management etc.) It FREQUENTLY spikes the CPU at 5%-10% – not an infrequent 2%.
“4) That depends. You don’t need 40 processes, but which ones are you going to turn off? Oh wait, you actually do need those. Damn. And more processes is better (better isolation of tasks, improves reliability and security).”
I run XP with only 16 processes running. It runs just fine and everything works, thanks. I’d need to enable a couple more to join a Domain, though.
“5) Why wouldn’t you want things to be cached in your cache (RAM is a cache)? If you don’t want things to be prefetched, just remove some RAM from your PC. Problem solved.”
Again. Dumbest statement ever. The only things that need to be pre-fetched are applications that *I* run regularly… not trying to pre-Fetch EVERYTHING (You seemed to have missed the key words “everything”). This is a major reason Vista shows taking up 500-750MB of RAM at desktop idle. WTF is it prefetching when I haven’t launched ANYTHING (Let alone even installed anything) yet? It’s certainly not something that *I WANT* pre-fetched.
“But you just asked not to prefetch things. This is a very confusing request. The OS uses more memory for applications, and you say that’s bad. Then you say what you want instead is for the OS to do what it’s already doing?”
How is hogging up more resources with Vista dedicating more resources to my applications, Einstein? If this OS is “so efficient”, why doesn’t it run well on a Celeron 500 with 256 MB of RAM, like I can get XP to do? Oh, forgive me… there is no “spoon”… err bloat.
The OS isn’t using more memory for applications… it’s using more memory for ITSELF… (see above response to your pre-fetching question).
“This is a perfect example of you not knowing what you’re asking for.
1) You cannot have a 32-bit OS that can use all 4GB of your physical memory. The two are mutually exclusive. 32-bit means 4GB of addresses. Unless the only thing you are addressing is physical memory, you will never be able to address 4GB of it. Since virtually all computers have other things to map addresses to (like, say, a video card) – it’s quite obvious that this is never going to happen.”
Ok… I flubbed up that statement… I’ll give you that. What I meant was zero need for a pagefile (On the Hard Disk) when using 4GB of RAM. I don’t care about crash dumps on my home computer… and am annoyed by Windows apps that require a pagefile no matter what (Like Photoshop). So I guess what I want is to be able to turn the paging executive off (Which I admit is possible) and force applications to deal with that scenerio. Also, windows only gives 2GB per process… If I’m running a process that can use that other Gig or so, I want it, leaving the final gig or so to the things you mentioned.
“That’s why we have 64-bit OSes now. If you have 4GB of memory, you should get one of those.”
I plan to, if one can run my entire software investment. If not, I’m stuck in 32-bit land (For Windows). My entire point of the rant was that Vista is WAY bloated. It’s not the typical 2x between Windows releases… it 4x – 8x in some areas. That’s just ridiculous. Windows 7 better be way cleaned up in comparison to Vista, or MS is going to lose even more people.
It’s not the kernel that’s the problem, it’s all the nontransparent crap stacked on top of it. That said though, kernels have evolved. I would like to see a truly hardware abstract kernel, running on a virtual cpu, giving binary cross platform portability. Oh and it would be so nice to see everything that belongs to a certain app/system to be contained in a single folder or package with shared components linked out virtually. EOL