Stony Stevenson alerts us to new information on the XP SP3-induced crashes that we discussed a few days back. Jesper Johansson, a former program manager for security policy at Microsoft, is maintaining an ongoing log and support site for users affected by any of several problems triggered by XP3. Machines using AMD hardware, particularly HP desktops, seem to have several modes of failure; others affect Intel machines.
Wow, you're right. With the Linux HP tablet market being about the same size as the Windows desktop + laptop market I'm surprised it didn't get more press.
Rewrite: "Over time, Microsoft Windows XP tends to mysteriously... decrease stability and performance (in every way imaginable".
I've experienced that, many times. Windows is unstable. The instability helps Microsoft sell new versions of its operating system.
Summary of the Slashdot article about Windows XP SP3 crashes:
Microsoft has known about one of the underlying problems for a long time. See KB888372 [microsoft.com]. It would have been easy to prevent the crashes merely by having SP3 installation do the work mentioned in the KB888372 article. However, apparently because of work avoidance, or an attempt to discourage people from using Windows XP, Microsoft did not do the necessary work.
my point was that SP3 was supposed to try to retain current windows users in the face of vista's failure to impress. SP3 instead seems to be driving people away at least until/if it gets fixed.
Quite simply, if MS wanted to keep customers they would create a product with zero problems (or as close as they can get) and push it out at a VERY competitive price. That is how the marketplace is supposed to work. When your namebrand is trashed, you have to compete extra hard. MS seems unwilling to do this, or at least has failed to show that they are trying to do so.
That might just be bad business decisions on their part, but whether it was malicious or stupidity does not matter. In either case the end result is that MS loses more customers. Nobody wanted to hear that MS was losing or soon to be dead a year ago when predictions were rife, but here it is, in your face. MS is consistently failing to either impress or produce quality product. The dragon^H^H^H^H^Hcathedral is near death... is it time for the penny market to celebrate?
Not on your life, it will be time to celebrate when the dried bones of the dragon are used up as party favors. Until then, it is time to keep competing aggressively, and nothing short of that will do. Competition, not patents, drives innovation. Innovation will bring us secure computing at home. A kind of secure that behaves friendly to the end user.
Now, am I bashing MS for pleasure? No, it is because MS products are in their deathbed and nothing short of a complete restart will get them out of it. It does not appear that MS will do that. There is nothing in current or near future activity that shows MS will do anything different from what got them in the death bed to start with. The beast is dieing. There is nothing more to say.
Call that a troll if you will, but the truth hurts sometimes. Do I want it to die? NO! Emphatically NO!!!! Without competition, quality dies. Would I like to see MS slide into a comfortable second place? Yes.... and the reasons are simple, just ask any Linux fanboi for them.
SP3 failed utterly in the face of the current market that MS faces. There is NO excuse for that in business. If you believe the art of war extends to business, MS deserves to be beheaded ungracefully. That is how business goes, so don't bother telling me that I'm a troll.
SP3 failed utterly in the face of the current market that MS faces.
While I agree with most of what you said, I'm not sure how SP3 "fails utterly". There have been far fewer problems with SP3 than previous service packs. Why, SP2, which is generally regarded as a Very Good Thing (and with good reason) broke a lot more software and machines than SP3 seems to be doing.
I'm usually the first to bash Microsoft when they deserve it, which is 99.9% of the time, but I cannot agree with the assessment that SP3 "fails".
Microsoft in general is "failing utterly" in the current market, but as far as I'm concerned, XP is doing just fine.
I'm guessing that you missed the news that if you are top dog, you can't afford to mess up?
It was not JUST tablet users. Read some news would you! IE8 beta users were screwed too. MS has had decades at working with EVERY kind of hardware. It's fscking lame to call that bluff now. F/OSS software might be able to still do that, but MS has NO excuse. period. for any reason. They have been working with this hardware FOR_EVER! I don't know how to say that strongly enough. Fuck! The hardware has been designed around the GD software. There is NO excuse. Business is business. Get it right or fail... this look like one more fail in the bag of fail that MS is filling up fairly fast. From a pure business pundit prospective, MS failed here. Keep drinking the coolaid!
Hopefully none. A smart business knows better than to run beta software for their mission-critical tasks.
A smart business would also know never to upgrade their systems the very moment an upgrade to a piece of software comes out. It's much smarter to wait a few weeks for the developers to figure out the problems that slipped by unnoticed during the beta stages due to fewer users. This is true for both free software and proprietary software. I remember having some nasty problems when I upgraded to the last two Ubuntu releases the day they came out. Now I'm waiting for Hardy to "stabilize" because I now know not to run software that's just been released. It's true that what we are talking about is just a service pack, but based on what happened when SP2 came out the public really should have expected Microsoft's future service packs to do just as much under-the-hood tinkering as SP2 did.
Yes, it is partially Microsoft's fault for not warning users on Automatic Updates that SP3 is still brand new and could potentially cause problems, but unless you never had problems with SP2 or were not in charge of a Windows XP machine during that time, this should have been seen from a mile away.
As a web developer, I routinely install browser betas so that I can catch any problems before they develop. I'm a business and that's called being proactive.
What I also do is run a zoo of Windows / Explorer combos in virtual machines. It's fun to do side by side comparisons. My summary: Vista is much slower at everything.
Headline singles out AMD machines, body indicates that AMD and Intel are equally affected by various modes of crash.
Sounds like someone's trying to drum down AMD stock or something... nah, we'd never have a processor partisan writing for Slash would we?
by Anonymous Coward
on Sunday May 11 2008, @11:46PM (#23374566)
You see, that's why Windows will never be ready for the desktop: Until you get rid of all that command line gibberish, I'll never be able to install it on my grandmother's computer.
No need for CLI in Windows, if you click "Hardware Settings", there is huge button flashing in many colours, saying "sc config intelppm start= disabled", just press it!
HP should NOT be using the same image for their Intel and AMD-based systems. There's always one for the Intel systems and one for AMD systems of each type (So, a DV2000 laptop has two generic system images, one for Intel-based and one for AMD-based. It's almost ALWAYS been this way.)
By the way, this appears to be Microsoft's problem, since HP maintains and is responsible for their own recovery images (all customized for each model and revision of laptop) and their own drivers.
HP should NOT be using the same image for their Intel and AMD-based systems.
In that case, I trust you'll forgive me asking this question:
What was the point in all the years spent by the PC industry on "Plug & Play", implementing ideas like unique IDs allocated by a manufacturer to their hardware devices and an operating system which can scan these IDs and choose drivers accordingly?
I guess I don't understand your POV. Just because you had to sit on it for work doesn't mean that's the right way to proceed. All these things have vendor strings and PCI IDs, Windows should be smart enough to ignore irrelevant drivers. Linux boots on all this stuff with a single kernel after all.
It's certainly Microsoft's fault that their operating system can't figure out at boot-time which drivers are appropriate for the platform it's booting on and only loading those.
Mac OS X and Linux both do this. Why can't Windows?
I installed SP3 Sunday and three problems immediately cropped up that I haven't seen in the years since I first installed XP. First is a stop, BAD_POOL_POINTER 0x00000019 (0x00000020,0x8a231120, 0x8a231158, 0x1a070000). Second is a problem with the HID service not starting. Third is that PaintShop Pro (V7) now cancels all attempt to enter standby mode. Sigh...
Too bad my submission from monday didn't make it, it would have made for some interesting conspiracy theories. AMD and Intel have made the briefs in their anti-trust case public (With heavy censorship^Wediting). One of AMD's contentions is that Intel's compiler is actually written to reduce speed and stability of programs it compiles when said programs are run on AMD processors.
<conspiracy>Maybe Microsoft has a deal with Intel to do the same with SP3 (and other Windows versions/SPs?) or they use Intel's compiler.</conspiracy>
more like intel payed HP to put there drivers on to all systems in way that MS says they do not support. HP needs to step up and pay for peoples down time and the cost of having a tech come out and fix it. This may even need to come down to a class action law suit.
make the head line say HP systems useing a unsupported by MS driver setup / image load crash under SP3.
It may also be worth pointing out that (according to TFA) folks had this exact same issue when service pack 2 came out, so it isn't as if HP's configuration wasn't already known to cause problems.
I imaged my whole Windows partition, in preparation for the horrible instability that would be SP3. I then took a deep breath, and started the download, figuring it would take several hours.
It went reasonably quickly, had exactly one reboot (which brought me fully up to date; no "critical updates" after that), and then ran solidly while I played Portal for another five or six hours.
Really: no-one has suggested that all machines have this problem after SP3 is installed, so one anecdote of a machine that does not suffer any problems is pointless.
As I understand it, this only impacts Windows XP users who are running computers with AMD or Intel processors. There is no evidence of SP3 introducing problems on XP machines with alternative architectures.
Just how many "alternative architectures" does XP run on? Last I checked, none. I think Microsoft's multi-architecture support for their main operating systems died after NT4 (along with support for DEC's Alpha) and they went x86. Looking at Microsoft's support page [microsoft.com], they say Pentium or compatible processor, so that means x86 only.
"SP3 causes the computer to crash during boot, and Windows XP, by default, is set up to automatically reboot when it crashes. That is why you end up in the endless rebooting scenario."
Nope, no relation at all. After all, crashing is perfectly normal.
I got hit by this bug when the patch went live last week on Windowsupdate. As the article states, the solution in was to disable intelppm.sys from safemode. It's a lot quicker if you do it using autoruns [microsoft.com]. It's too bad this article wasn't posted last week. It would have saved me a lot of trouble shooting time.
Anyone who build Windows XP images that are rolled out onto both AMD and Intel machines should have long ago learned about the stop code 0x0000007E perils that come from the intelppm driver. The root of all evil here is that processors are not plug and play devices as far as XP is concerned and their associated drivers are hardcoded to start at boot time. Why the hell Microsoft has not taken the time to update intelppm.sys to check for a GenuineIntel x86 Family XX Model YY Stepping ZZ ID before touching HW specific registers is a mystery to me (I hope the conspiricy theorists amongst you will regale me with much food for thought).
That's not the only issue with SP3. One of my monitors is rotated 90 degrees (widescreen that I use upright), thanks to the ATI driver's rotate function.
After rebooting following SP3 install, all my monitors went completely berzerk. They fell back to 4 bits colors (I didn't even know there WAS a 4 bit mode), with some weird effects. Also, rotation was not possible.
It took me about an hour to find a way to bring back monitors to decent resolution and colors. I still couldn't get rotation to work, no matter how hard I tried (Combination of card, drivers, update from ATI, etc)
Then finally I google a bit and found a few forums with user complaints of the same type of problem. So I uninstalled SP3, rebooted, and voilà, everything back to normal.
Needless to say, I promply logged back into WSUS and removed SP3 from the approved for installed list.
Only had one problem on my Acer Laptop (TravelMate 8210) with SP3, it forgot that I had a wireless card, restarted and been fine since then (almost a week) so I'd say no problems for me
Wow! Your anecdote (in which you don't even mention if you're using intel or AMD) has totally changed my mind about the reliability of SP3!
Typical Microsoft (Score:4, Funny)
Re:Typical Microsoft (Score:5, Funny)
Parent
Re:Typical Microsoft (Score:5, Funny)
Parent
Re:Typical Microsoft (Score:5, Insightful)
Parent
Microsoft knowingly released unfinished software? (Score:5, Informative)
I've experienced that, many times. Windows is unstable. The instability helps Microsoft sell new versions of its operating system.
Summary of the Slashdot article about Windows XP SP3 crashes:
Microsoft has known about one of the underlying problems for a long time. See KB888372 [microsoft.com]. It would have been easy to prevent the crashes merely by having SP3 installation do the work mentioned in the KB888372 article. However, apparently because of work avoidance, or an attempt to discourage people from using Windows XP, Microsoft did not do the necessary work.
Parent
Ulterior motive (Score:5, Funny)
Re:Ulterior motive (Score:4, Insightful)
Parent
Re: (Score:3, Insightful)
Re:Ulterior motive (Score:4, Insightful)
Parent
There is only one problem with this theory (Score:5, Insightful)
That might just be bad business decisions on their part, but whether it was malicious or stupidity does not matter. In either case the end result is that MS loses more customers. Nobody wanted to hear that MS was losing or soon to be dead a year ago when predictions were rife, but here it is, in your face. MS is consistently failing to either impress or produce quality product. The dragon^H^H^H^H^Hcathedral is near death... is it time for the penny market to celebrate?
Not on your life, it will be time to celebrate when the dried bones of the dragon are used up as party favors. Until then, it is time to keep competing aggressively, and nothing short of that will do. Competition, not patents, drives innovation. Innovation will bring us secure computing at home. A kind of secure that behaves friendly to the end user.
Now, am I bashing MS for pleasure? No, it is because MS products are in their deathbed and nothing short of a complete restart will get them out of it. It does not appear that MS will do that. There is nothing in current or near future activity that shows MS will do anything different from what got them in the death bed to start with. The beast is dieing. There is nothing more to say.
Call that a troll if you will, but the truth hurts sometimes. Do I want it to die? NO! Emphatically NO!!!! Without competition, quality dies. Would I like to see MS slide into a comfortable second place? Yes.... and the reasons are simple, just ask any Linux fanboi for them.
SP3 failed utterly in the face of the current market that MS faces. There is NO excuse for that in business. If you believe the art of war extends to business, MS deserves to be beheaded ungracefully. That is how business goes, so don't bother telling me that I'm a troll.
Parent
Re:There is only one problem with this theory (Score:5, Interesting)
While I agree with most of what you said, I'm not sure how SP3 "fails utterly". There have been far fewer problems with SP3 than previous service packs. Why, SP2, which is generally regarded as a Very Good Thing (and with good reason) broke a lot more software and machines than SP3 seems to be doing.
I'm usually the first to bash Microsoft when they deserve it, which is 99.9% of the time, but I cannot agree with the assessment that SP3 "fails".
Microsoft in general is "failing utterly" in the current market, but as far as I'm concerned, XP is doing just fine.
Parent
Re:There is only one problem with this theory (Score:4, Insightful)
It was not JUST tablet users. Read some news would you! IE8 beta users were screwed too. MS has had decades at working with EVERY kind of hardware. It's fscking lame to call that bluff now. F/OSS software might be able to still do that, but MS has NO excuse. period. for any reason. They have been working with this hardware FOR_EVER! I don't know how to say that strongly enough. Fuck! The hardware has been designed around the GD software. There is NO excuse. Business is business. Get it right or fail... this look like one more fail in the bag of fail that MS is filling up fairly fast. From a pure business pundit prospective, MS failed here. Keep drinking the coolaid!
Parent
Re:There is only one problem with this theory (Score:4, Insightful)
Hopefully none. A smart business knows better than to run beta software for their mission-critical tasks.
A smart business would also know never to upgrade their systems the very moment an upgrade to a piece of software comes out. It's much smarter to wait a few weeks for the developers to figure out the problems that slipped by unnoticed during the beta stages due to fewer users. This is true for both free software and proprietary software. I remember having some nasty problems when I upgraded to the last two Ubuntu releases the day they came out. Now I'm waiting for Hardy to "stabilize" because I now know not to run software that's just been released. It's true that what we are talking about is just a service pack, but based on what happened when SP2 came out the public really should have expected Microsoft's future service packs to do just as much under-the-hood tinkering as SP2 did.
Yes, it is partially Microsoft's fault for not warning users on Automatic Updates that SP3 is still brand new and could potentially cause problems, but unless you never had problems with SP2 or were not in charge of a Windows XP machine during that time, this should have been seen from a mile away.
Parent
Re:There is only one problem with this theory (Score:4, Interesting)
As a web developer, I routinely install browser betas so that I can catch any problems before they develop. I'm a business and that's called being proactive.
What I also do is run a zoo of Windows / Explorer combos in virtual machines. It's fun to do side by side comparisons. My summary: Vista is much slower at everything.
Parent
Re:Ulterior motive (Score:5, Funny)
Parent
Hey, wait a minute! (Score:3, Insightful)
Re:Hey, wait a minute! (Score:5, Informative)
Easiest way to fix the problem, before installing SP3, open a CMD window, and type "sc config intelppm start= disabled".
Parent
Re:Hey, wait a minute! (Score:5, Funny)
Parent
Re:Hey, wait a minute! (Score:4, Funny)
Parent
Re:Hey, wait a minute! (Score:5, Informative)
http://msinfluentials.com/blogs/jesper/archive/2008/05/08/does-your-amd-based-computer-boot-after-installing-xp-sp3.aspx [msinfluentials.com]
Johansson (partly) blames HP and other OEMs for using the same disk image for Intel and AMD PCs. He also gives some directions on fixing the problems.
Parent
Rename the topic to say INTEL drivers on AMD systm (Score:5, Informative)
The topic you have makes AMD look bad.
Why is HP useing the same basic image for there amd and intel systems?
What other driver bloat is in OEM systems?
Is INTEL coding there drivers to mess up AMD systems?
AMD legal should take a look at this.
I have SP3 running on my AMD right now and it's works 100%
Re:Rename the topic to say INTEL drivers on AMD sy (Score:5, Interesting)
By the way, this appears to be Microsoft's problem, since HP maintains and is responsible for their own recovery images (all customized for each model and revision of laptop) and their own drivers.
Parent
Re:Rename the topic to say INTEL drivers on AMD sy (Score:4, Insightful)
What was the point in all the years spent by the PC industry on "Plug & Play", implementing ideas like unique IDs allocated by a manufacturer to their hardware devices and an operating system which can scan these IDs and choose drivers accordingly?
Parent
Re:Rename the topic to say INTEL drivers on AMD sy (Score:5, Insightful)
Parent
Re:Rename the topic to say INTEL drivers on AMD sy (Score:4, Insightful)
Parent
Re:Rename the topic to say INTEL drivers on AMD sy (Score:4, Insightful)
Mac OS X and Linux both do this. Why can't Windows?
Parent
BAD_POOL_POINTER, HID problem, standyby problem (Score:5, Informative)
Wintel Conspiracy (Score:4, Interesting)
<conspiracy>Maybe Microsoft has a deal with Intel to do the same with SP3 (and other Windows versions/SPs?) or they use Intel's compiler.</conspiracy>
Worth considering.
Re:Wintel Conspiracy (Score:5, Insightful)
make the head line say HP systems useing a unsupported by MS driver setup / image load crash under SP3.
Parent
Re:Wintel Conspiracy (Score:5, Informative)
Parent
Ancedote time... (Score:5, Funny)
It went reasonably quickly, had exactly one reboot (which brought me fully up to date; no "critical updates" after that), and then ran solidly while I played Portal for another five or six hours.
I was almost disappointed.
It was an Intel machine, though.
Re: (Score:3, Interesting)
Really: no-one has suggested that all machines have this problem after SP3 is installed, so one anecdote of a machine that does not suffer any problems is pointless.
Limited impact (Score:5, Funny)
Re:Limited impact (Score:4, Funny)
Thats what all you fanboy AMD and Intel people deserve.
Parent
Re: (Score:3, Informative)
misleading headline (Score:5, Insightful)
not exactly a cut and dry SP3 problem and certainly not an AMD or INTEL issue at all.
people who write this crap need to all be thrown in a cage and be made to rip each other apart.
endless rebooting is not at all related to SP3 per (Score:3, Funny)
Nope, no relation at all. After all, crashing is perfectly normal.
Im waiting.. (Score:5, Funny)
OK so far (Score:3, Informative)
Microsoft: (Score:5, Funny)
The World is our Beta Tester.
Blue screen after first reboot... (Score:5, Informative)
Stop code 0x0000007E is not a new problem (Score:5, Informative)
Screen rotation (Score:5, Informative)
After rebooting following SP3 install, all my monitors went completely berzerk. They fell back to 4 bits colors (I didn't even know there WAS a 4 bit mode), with some weird effects. Also, rotation was not possible.
It took me about an hour to find a way to bring back monitors to decent resolution and colors. I still couldn't get rotation to work, no matter how hard I tried (Combination of card, drivers, update from ATI, etc)
Then finally I google a bit and found a few forums with user complaints of the same type of problem. So I uninstalled SP3, rebooted, and voilà, everything back to normal.
Needless to say, I promply logged back into WSUS and removed SP3 from the approved for installed list.
Bullshit! (Score:4, Funny)
Re:Frist Pr0st (Score:5, Funny)
Parent
Re:Frist Pr0st (Score:4, Funny)
Uhh, hello? Anybody still there?
Parent
Re:Frist Pr0st (Score:5, Funny)
Parent
Re:Only one crash (Score:4, Funny)
Wow! Your anecdote (in which you don't even mention if you're using intel or AMD) has totally changed my mind about the reliability of SP3!
Parent
Re:Only one crash (Score:5, Funny)
Parent
Re:Only one crash (Score:5, Funny)
Parent