"Most Stable OS ever": Lie, or mistaken?
-
The basic problem is that people seem to think that it's the OS's fault if users install bad drivers. If a driver couldn't crash your system, I'd think it would also not have the sort of access to it that it needs to work. I've had one install of W2000 die, in that IE stopped working and a few other things, but apart from that one I have always found it very reliable. In fact, between us here at work, I've seen it blue screen something like four times, and when it happened we told each other because it was such a rare event. We have all ditched W98 as being very liable to die in comparison. I ran NT 4.0 as a development platform for non-graphics work prior and would agree that W2000 was the most reliable Windows release ever, I say was because I'm yet to crash XP, and haven't used it long enough to comment between the two. The thread you pointed to relates to something that has been discussed here at length in the last couple of days. I don't think finding one bug that can cause a specific line of code ( that no-one is likely to write in the 'real world' ) executed over and over again ad infinitum to bring the system down makes the OS unstable. Christian After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
Wow. Lots of apologists, no calls for accountability. Yes, clearly the bug is in the driver, in the first case. The second case is a simple, legitimate test though. But the bottom line is that this "most stable OS ever" still crashes, and crashes with a great big boom. Engineers are funny at times - as long as we understand why things go wrong, we are sometimes forgiving of that; forgetting that the average user does not give a damn about reasons. It's possible to crash Windows 2K. It's easy to crash Windows 2K. Windows 2K can crash out of the box. And why do I continue to point the finger at MS, even for the video driver? Because they continue to make development harder, more complex, and more reliant on their fragile "component" philosophy. Device driver developers have had to contend with a multitude of different driver models over the past few years. With product life cycles measured in months, a new driver model to contend with every few years.. no wonder bugs happen. Time to get back to work..
-
Wow. Lots of apologists, no calls for accountability. Yes, clearly the bug is in the driver, in the first case. The second case is a simple, legitimate test though. But the bottom line is that this "most stable OS ever" still crashes, and crashes with a great big boom. Engineers are funny at times - as long as we understand why things go wrong, we are sometimes forgiving of that; forgetting that the average user does not give a damn about reasons. It's possible to crash Windows 2K. It's easy to crash Windows 2K. Windows 2K can crash out of the box. And why do I continue to point the finger at MS, even for the video driver? Because they continue to make development harder, more complex, and more reliant on their fragile "component" philosophy. Device driver developers have had to contend with a multitude of different driver models over the past few years. With product life cycles measured in months, a new driver model to contend with every few years.. no wonder bugs happen. Time to get back to work..
1/ Who is accountable, Microsoft or the person who wrote your bodgy driver ? The answer is simple, take a quick poll of what percentage of W2000 users have had this sort of problem, and if the problem can be traced back to specific drivers. 2/ A legitimate test ? I grant you that it's strange that it kills the OS, but what is it a legitimate test of, except a test of if this line crashes the system ? It's not like it's something anyone is ever likely to *do*. 3/ It's easy to crash a Volvo, does that mean they don't deserve the title 'safest car on the road' ( which I *believe* they hold, if not, insert the appropriate manufacturer ). The *most stable* title implies a degree of instability, otherwise it would be 'completely stable'. Win2K can crash out of the box if you compile and run a snippet that is not ever going to be executed in any production code, or if you're installing badly written drivers. That does not stop it being the *most stable*, indeed that statement makes only promises relative to other OS's and by itself says close to nothing. 4/ I've never written a driver myself, so I can't comment specifically, but I don't believe M$ are making development any harder at all. Christian After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
-
Wow. Lots of apologists, no calls for accountability. Yes, clearly the bug is in the driver, in the first case. The second case is a simple, legitimate test though. But the bottom line is that this "most stable OS ever" still crashes, and crashes with a great big boom. Engineers are funny at times - as long as we understand why things go wrong, we are sometimes forgiving of that; forgetting that the average user does not give a damn about reasons. It's possible to crash Windows 2K. It's easy to crash Windows 2K. Windows 2K can crash out of the box. And why do I continue to point the finger at MS, even for the video driver? Because they continue to make development harder, more complex, and more reliant on their fragile "component" philosophy. Device driver developers have had to contend with a multitude of different driver models over the past few years. With product life cycles measured in months, a new driver model to contend with every few years.. no wonder bugs happen. Time to get back to work..
It's easy to crash Windows 2K Okay, four lines of code, but as a programmer who has used C++ for years I didn't know this was possible, and certainly nobody in the non-programming profession will. But the bottom line is that this "most stable OS ever" still crashes Nobody said it never crashes, just that it doesn't crash as often as previous versions. People seem to think that the *nix OSes are pure stablility itself but they aren't - they crash too. Just some food for thought :) -- Andrew.
-
Wow. Lots of apologists, no calls for accountability. Yes, clearly the bug is in the driver, in the first case. The second case is a simple, legitimate test though. But the bottom line is that this "most stable OS ever" still crashes, and crashes with a great big boom. Engineers are funny at times - as long as we understand why things go wrong, we are sometimes forgiving of that; forgetting that the average user does not give a damn about reasons. It's possible to crash Windows 2K. It's easy to crash Windows 2K. Windows 2K can crash out of the box. And why do I continue to point the finger at MS, even for the video driver? Because they continue to make development harder, more complex, and more reliant on their fragile "component" philosophy. Device driver developers have had to contend with a multitude of different driver models over the past few years. With product life cycles measured in months, a new driver model to contend with every few years.. no wonder bugs happen. Time to get back to work..
I understand all the issues of drivers, testing, etc. All I can say is that I've had Win2000 installed as my work operating system - I use it 7 days a week, it runs 24 hours a day (I never turn the thing off). When I'm in front of it I run Outlook, Explorer, IE, CodeCo-op and VC permanently open. I often open as many as 5 or 6 IE windows simultaneously. I debug multiple programs simultaneously. I often start a full build, start a download from the net, send some files to the network, and do a search of the MSDN DVD disk all at the same time. The programs I write and debug crash regularly (while I'm writing them, not after I finish!!), so I have programs permanently crashing on a regular basis, with all that that implies - dangling DLLs, GDI resource leaks, etc. When I'm not in front of the machine it runs the 'seti@home' client (I've logged over 3000 hours of seti so far), plus Outlook. I reboot the machine only when I need to (normally after installing some software, or a 'windows upgrade') - this normally means that I go anywhere up to 6 or 7 weeks without a reboot. In the 6 plus months I've been running this, it has crashed exactly once. So, other people may have different experiences of course, but I have to say that win2000 could not be more stable in my environment. Testimonials are the worst form of evidence, but all I can say is that I have felt no regrets in making win2000 the basis for my work environment. ----------------------- The sermon on the mount... Man 1 : Hear that? Blessed are the greek. Man 2 : The greek? Man 1 : Well apparently, he's going to inherit the earth. Man 2 : Did anyone catch his name?
-
Ever since I got my first Win2K system, which spontaneously rebooted frequently due to a buggy video driver, I've been > beyond< skeptical of MS's claim that this version of NT is their "Most Stable OS ever". Prior to this I hade never had one of MS's OS's crash so hard (it would frequently trash files I was working on in VC++). The following thread on Usenet is even more damning: http://groups.google.com/groups?hl=en&threadm=3bd56bdb%240%2421707%2444c9b20d%40news2.asahi-net.or.jp&rnum=1&prev=/groups%3Fq%3DBug%2Bwhich%2Bcrash%2BWindows%2B2000%26hl%3Den%26rnum%3D1%26selm%3D3bd56bdb%25240%252421707%252444c9b20d%2540news2.asahi-net.or.jp My question is: When MS makes statements like this, are they simply lying? Or is their testing so bad that they can't tell how unstable their code is?
-
Wow. Lots of apologists, no calls for accountability. Yes, clearly the bug is in the driver, in the first case. The second case is a simple, legitimate test though. But the bottom line is that this "most stable OS ever" still crashes, and crashes with a great big boom. Engineers are funny at times - as long as we understand why things go wrong, we are sometimes forgiving of that; forgetting that the average user does not give a damn about reasons. It's possible to crash Windows 2K. It's easy to crash Windows 2K. Windows 2K can crash out of the box. And why do I continue to point the finger at MS, even for the video driver? Because they continue to make development harder, more complex, and more reliant on their fragile "component" philosophy. Device driver developers have had to contend with a multitude of different driver models over the past few years. With product life cycles measured in months, a new driver model to contend with every few years.. no wonder bugs happen. Time to get back to work..
Oh Boo Hoo, Of course it can crash, if you intentionally f2k with it with the express purpose of bringing it down. I'd go so far as saying that you could probably do that to any OS if you spent enough time or effort in doing so. Having said that. Why the hell would you want to? It seems to me that the only people trying to make Windows crash, and then whinge about it, are those that have something against MS/Windows in the first place. As for stability. I've had W2K running for around 18 months. The only BSODs I had were on my old system during its terminal illness. A note here W2K spontaneously rebooted when running Java programs mostly, Win98 didnt even boot up. However, it was my motherboard/CPU that was at fault here. Since I've had W2K on my new box, it has worked perfectly while left running for weeks at a time. The only probs I have had it when 3RD PARTY drivers dont deallocate memory and I have to reboot. Very rare indeed. Maybe I am an apologist for MS, I dont give a damn. I like the software and find it suits my purposes with a minimum of fuss.
-
It's worth noting that NT4 did this too. This was (as I understand) because many video functions were brought into the kernel. They were brought into the kernel to improve graphics performance. Craig Dodge A catchy signature should appear here.
A catchy signature should appear here. Hey, let us know when you've thought of one...or failing that ask John for ideas. I'm sure he'll have a couple of suggestions... :eek: Andy Metcalfe - Sonardyne International Ltd
(andy.metcalfe@lineone.net)
http://www.resorg.co.uk"I'm just another 'S' bend in the internet. A ton of stuff goes through my system, and some of the hairer, stickier and lumpier stuff sticks." - Chris Maunder (I just couldn't let that one past ;))
-
I understand all the issues of drivers, testing, etc. All I can say is that I've had Win2000 installed as my work operating system - I use it 7 days a week, it runs 24 hours a day (I never turn the thing off). When I'm in front of it I run Outlook, Explorer, IE, CodeCo-op and VC permanently open. I often open as many as 5 or 6 IE windows simultaneously. I debug multiple programs simultaneously. I often start a full build, start a download from the net, send some files to the network, and do a search of the MSDN DVD disk all at the same time. The programs I write and debug crash regularly (while I'm writing them, not after I finish!!), so I have programs permanently crashing on a regular basis, with all that that implies - dangling DLLs, GDI resource leaks, etc. When I'm not in front of the machine it runs the 'seti@home' client (I've logged over 3000 hours of seti so far), plus Outlook. I reboot the machine only when I need to (normally after installing some software, or a 'windows upgrade') - this normally means that I go anywhere up to 6 or 7 weeks without a reboot. In the 6 plus months I've been running this, it has crashed exactly once. So, other people may have different experiences of course, but I have to say that win2000 could not be more stable in my environment. Testimonials are the worst form of evidence, but all I can say is that I have felt no regrets in making win2000 the basis for my work environment. ----------------------- The sermon on the mount... Man 1 : Hear that? Blessed are the greek. Man 2 : The greek? Man 1 : Well apparently, he's going to inherit the earth. Man 2 : Did anyone catch his name?
I think this speaks for itself. Well spoken Mike. :) Andy Metcalfe - Sonardyne International Ltd
(andy.metcalfe@lineone.net)
http://www.resorg.co.uk"I'm just another 'S' bend in the internet. A ton of stuff goes through my system, and some of the hairer, stickier and lumpier stuff sticks." - Chris Maunder (I just couldn't let that one past ;))
-
A catchy signature should appear here. Hey, let us know when you've thought of one...or failing that ask John for ideas. I'm sure he'll have a couple of suggestions... :eek: Andy Metcalfe - Sonardyne International Ltd
(andy.metcalfe@lineone.net)
http://www.resorg.co.uk"I'm just another 'S' bend in the internet. A ton of stuff goes through my system, and some of the hairer, stickier and lumpier stuff sticks." - Chris Maunder (I just couldn't let that one past ;))
Hmmm...somebody's deleted a post here methinks. :rolleyes: Andy Metcalfe - Sonardyne International Ltd
(andy.metcalfe@lineone.net)
http://www.resorg.co.uk"I'm just another 'S' bend in the internet. A ton of stuff goes through my system, and some of the hairer, stickier and lumpier stuff sticks." - Chris Maunder (I just couldn't let that one past ;))
-
A catchy signature should appear here. Hey, let us know when you've thought of one...or failing that ask John for ideas. I'm sure he'll have a couple of suggestions... :eek: Andy Metcalfe - Sonardyne International Ltd
(andy.metcalfe@lineone.net)
http://www.resorg.co.uk"I'm just another 'S' bend in the internet. A ton of stuff goes through my system, and some of the hairer, stickier and lumpier stuff sticks." - Chris Maunder (I just couldn't let that one past ;))
>> A catchy signature should appear here. > > Hey, let us know when you've thought of one...or failing that ask John for ideas. I'm sure he'll > have a couple of suggestions... I already provide plenty of quotable material. He just has to scan back through the lounge or the rant/rave forum and pick the one he likes. :) To hell with those thin-skinned pillow-biters. - Me, 10/03/2001
-
I understand all the issues of drivers, testing, etc. All I can say is that I've had Win2000 installed as my work operating system - I use it 7 days a week, it runs 24 hours a day (I never turn the thing off). When I'm in front of it I run Outlook, Explorer, IE, CodeCo-op and VC permanently open. I often open as many as 5 or 6 IE windows simultaneously. I debug multiple programs simultaneously. I often start a full build, start a download from the net, send some files to the network, and do a search of the MSDN DVD disk all at the same time. The programs I write and debug crash regularly (while I'm writing them, not after I finish!!), so I have programs permanently crashing on a regular basis, with all that that implies - dangling DLLs, GDI resource leaks, etc. When I'm not in front of the machine it runs the 'seti@home' client (I've logged over 3000 hours of seti so far), plus Outlook. I reboot the machine only when I need to (normally after installing some software, or a 'windows upgrade') - this normally means that I go anywhere up to 6 or 7 weeks without a reboot. In the 6 plus months I've been running this, it has crashed exactly once. So, other people may have different experiences of course, but I have to say that win2000 could not be more stable in my environment. Testimonials are the worst form of evidence, but all I can say is that I have felt no regrets in making win2000 the basis for my work environment. ----------------------- The sermon on the mount... Man 1 : Hear that? Blessed are the greek. Man 2 : The greek? Man 1 : Well apparently, he's going to inherit the earth. Man 2 : Did anyone catch his name?
I can crash Windows 2000 to the point where there's not even a BSOD. It just all of a sudden reboots. Any programmer worth his salt can do this through... It's more stable than anything that came before it though. To hell with those thin-skinned pillow-biters. - Me, 10/03/2001
-
1/ Who is accountable, Microsoft or the person who wrote your bodgy driver ? The answer is simple, take a quick poll of what percentage of W2000 users have had this sort of problem, and if the problem can be traced back to specific drivers. 2/ A legitimate test ? I grant you that it's strange that it kills the OS, but what is it a legitimate test of, except a test of if this line crashes the system ? It's not like it's something anyone is ever likely to *do*. 3/ It's easy to crash a Volvo, does that mean they don't deserve the title 'safest car on the road' ( which I *believe* they hold, if not, insert the appropriate manufacturer ). The *most stable* title implies a degree of instability, otherwise it would be 'completely stable'. Win2K can crash out of the box if you compile and run a snippet that is not ever going to be executed in any production code, or if you're installing badly written drivers. That does not stop it being the *most stable*, indeed that statement makes only promises relative to other OS's and by itself says close to nothing. 4/ I've never written a driver myself, so I can't comment specifically, but I don't believe M$ are making development any harder at all. Christian After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
Why does anyone get so engaged when s.o. has the opinion that MS sucks? Especially when this s.o. has something to prove that :) Honestly, in our company with lots of computers i tried it on 4 quite different ones (laptop, server (w2k server), developer machine (w2k prof) and an older developer machine (nt4 prof)) ... Guess what? Crashed on any of those. This makes it IMHO kind of implausible that this is a driver problem, because the driver structure has been changed completely from NT4 to W2k - and yes, christian, your believing is wrong in point 4) - just take a look at the topics WDM (Me), WDM (2k), DDK NT, DDK 2k, DDK 98+, ...
-
I can crash Windows 2000 to the point where there's not even a BSOD. It just all of a sudden reboots. Any programmer worth his salt can do this through... It's more stable than anything that came before it though. To hell with those thin-skinned pillow-biters. - Me, 10/03/2001
I can crash Windows 2000 to the point where there's not even a BSOD. It just all of a sudden reboots I'm sure that win200 crashes - I just wanted to point out that for some unforseen reason Billy-Bob Gates seems to have given me an unusually stable version, and this makes me very happy! I started in the windows world back with "windows 286", so I find win2000 to be a very acceptable desktop environment. Any programmer worth his salt can do this through... I mustn't be trying hard enough! Leave it with me ... but is this really a skill I need ?? ----------------------------- The sermon on the mount... Man 1 : Hear that? Blessed are the greek. Man 2 : The greek? Man 1 : Well apparently, he's going to inherit the earth. Man 2 : Did anyone catch his name? ----------------------------- "Happy pooja to all curry munchers!" - Paul Watson Oct 25, 2001
-
Ever since I got my first Win2K system, which spontaneously rebooted frequently due to a buggy video driver, I've been > beyond< skeptical of MS's claim that this version of NT is their "Most Stable OS ever". Prior to this I hade never had one of MS's OS's crash so hard (it would frequently trash files I was working on in VC++). The following thread on Usenet is even more damning: http://groups.google.com/groups?hl=en&threadm=3bd56bdb%240%2421707%2444c9b20d%40news2.asahi-net.or.jp&rnum=1&prev=/groups%3Fq%3DBug%2Bwhich%2Bcrash%2BWindows%2B2000%26hl%3Den%26rnum%3D1%26selm%3D3bd56bdb%25240%252421707%252444c9b20d%2540news2.asahi-net.or.jp My question is: When MS makes statements like this, are they simply lying? Or is their testing so bad that they can't tell how unstable their code is?
-
The basic problem is that people seem to think that it's the OS's fault if users install bad drivers. If a driver couldn't crash your system, I'd think it would also not have the sort of access to it that it needs to work. I've had one install of W2000 die, in that IE stopped working and a few other things, but apart from that one I have always found it very reliable. In fact, between us here at work, I've seen it blue screen something like four times, and when it happened we told each other because it was such a rare event. We have all ditched W98 as being very liable to die in comparison. I ran NT 4.0 as a development platform for non-graphics work prior and would agree that W2000 was the most reliable Windows release ever, I say was because I'm yet to crash XP, and haven't used it long enough to comment between the two. The thread you pointed to relates to something that has been discussed here at length in the last couple of days. I don't think finding one bug that can cause a specific line of code ( that no-one is likely to write in the 'real world' ) executed over and over again ad infinitum to bring the system down makes the OS unstable. Christian After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
The sample program that someone posted wasn't a stress problem, it was an actual bug in the OS. I ran the program in the debugger and it crashes the OS the second time it executes the print. Really stupid bug. [slaps MS around some] Tim Smith Descartes Systems Sciences, Inc.
-
Why does anyone get so engaged when s.o. has the opinion that MS sucks? Especially when this s.o. has something to prove that :) Honestly, in our company with lots of computers i tried it on 4 quite different ones (laptop, server (w2k server), developer machine (w2k prof) and an older developer machine (nt4 prof)) ... Guess what? Crashed on any of those. This makes it IMHO kind of implausible that this is a driver problem, because the driver structure has been changed completely from NT4 to W2k - and yes, christian, your believing is wrong in point 4) - just take a look at the topics WDM (Me), WDM (2k), DDK NT, DDK 2k, DDK 98+, ...
Do you mean the \b\b thing ? I never said that that was a driver problem, I said it was obviously a specific bug that nevertheless cannot be extrapolated to mean the rest of the OS is unstable, nor is it a line of code likely to ever be run in the real world. It's a bug, it crashes W2000 ( in fact the first time, the loop is not needed, I played with it today too ), but that does not mean there are other similar bugs, nor is that bug going to cause instability amongst people who don't compile that line of code for themselves to see what it does. 4/ I've never written a driver myself, so I can't comment specifically, but I don't believe M$ are making development any harder at all. I specifically said I do not write drivers, are you blind ? How can the driver development *possibly* stay the same if the OS changes ? I stand by my point, by and large I vehemently disagree that M$ is making development harder - they even have a new language to make it easier for monkeys to write programs. Christian After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
-
The sample program that someone posted wasn't a stress problem, it was an actual bug in the OS. I ran the program in the debugger and it crashes the OS the second time it executes the print. Really stupid bug. [slaps MS around some] Tim Smith Descartes Systems Sciences, Inc.
Actually I played with it today and it appears to crash the first time, and what happens is if you have some tabs, and then an equal number of backspaces (\t\t\b\b) then the second \b also deletes all the text before it, and a third \b will kill the OS, I'm guessing by stepping back too far. It's the combination of \t and ]b, just printing 'test\b\b\b\b\b' won't crash it, but add a \t before the \b's and you're toast. I never meant to imply it wasn't a bug in the OS, only that the existence of such a bug does not make the whole OS unstable. Christian After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
-
Wow. Lots of apologists, no calls for accountability. Yes, clearly the bug is in the driver, in the first case. The second case is a simple, legitimate test though. But the bottom line is that this "most stable OS ever" still crashes, and crashes with a great big boom. Engineers are funny at times - as long as we understand why things go wrong, we are sometimes forgiving of that; forgetting that the average user does not give a damn about reasons. It's possible to crash Windows 2K. It's easy to crash Windows 2K. Windows 2K can crash out of the box. And why do I continue to point the finger at MS, even for the video driver? Because they continue to make development harder, more complex, and more reliant on their fragile "component" philosophy. Device driver developers have had to contend with a multitude of different driver models over the past few years. With product life cycles measured in months, a new driver model to contend with every few years.. no wonder bugs happen. Time to get back to work..
Wow. Lots of apologists, no calls for accountability. No Jim, it isn't about being an apologist or not calling for MS to be accountable, it is about living in the real world. 1. Operating systems (all) have bugs. 2. Microsoft will fix the problem. So get off your stupid MS is evil high horse and get a clue about what the computer world is really all about. God, I am getting sick of all this crap. Just more mindless junk coming from people who have no real clue. Tim Smith Descartes Systems Sciences, Inc.
-
>> A catchy signature should appear here. > > Hey, let us know when you've thought of one...or failing that ask John for ideas. I'm sure he'll > have a couple of suggestions... I already provide plenty of quotable material. He just has to scan back through the lounge or the rant/rave forum and pick the one he likes. :) To hell with those thin-skinned pillow-biters. - Me, 10/03/2001
My current favorite is: The reason I asked is because I want to make sure I target the more correct description of your mental deficiency before I start making fun of it.