Some processes like servlet containers and their servlets could, in theory, be moved entirely into kernel land, without having to program any kernel code at all.
What the fuck? You actually think that's a good idea?
Eh. Think back to the blog about the start menu->shut down menu thing that took a year to develop. Communication between groups in Windows is not very good, and so when the Vista guys fix a vulnerability there, the XP guys don't necessarily get the news. It's not exactly a redeeming point, I know, but I doubt there's anything malicious or intentional going on. Like most things, it's just disorganization and red tape.
I agree that the drivers are at fault; this reeks of a mistake on Intel's part inside the GMA driver. All of the video card manufacturers have been going through hell working on getting things working smoothly in the new driver model. While Apple engineers are generally fairly incompetent at doing actual Windows coding, I'm firmly in the camp that an application should never be able to cause a BSOD.
Well. Let me show you a screenshot I snapped early on in the IE betas. I think this was beta 1, but it's been a while now. And just to be clear, I took the shot myself. It's not doctored in any way. It happened while I was on the MS corporate network, and you can even see that the Passport login passed.
Flagged for Phishing
The thing is, the e-mail doesn't say what the (quite obviously biased) macworld claims it says. That's the beauty of selective quoting. Reading the rest of the message gives a somewhat different perspective.
The pace of our discussions with Apple as well as their recent unsatisfactory response have certainly
frustrated a tot of people at Microsoft. The threat to Cancel Mac Office 97 is certainly the strongest bar9aining
pointwe have, as doing so wil do a great deal of harm to Apple immedIately. I also believe that Apple is
taking this threat pretty seriously, and at least someone there seems to want to move forward (when I
discussed the Issue wfth Jim Gable, an Apple marketing VPwho visited MS today, he seemed very cortcamed
aboutgetting more details on our specificobjections to their latest proposal; also, we received mail today
from Apple's evangelism group asking for details on the Office Early Mopter Program, saying that exec mgmt
had instructed them to get these detaIls (participating in this program was one of the minor issues in the
discussions)).
Regardless of the outcome of these discussions, though, I believe weshould ship Mac Office 97
Furfhermore, I believe we need to decide this immediately - our indecision so far has caused quite a bit of
harm, and this will become farworse very shortly, as we are not only close to shipping code externally, but
need to finally start press and customer communications, especially with MacWorld a month away.
Later on in the email, we see some perspective on what exactly the "testing features" were:
Because Mac Office Is so much less critical to our business than windows, we have the flexibility to test
out new things in the product and in its marketing before we try them~onWindows. Setup-less install, for
example, is one thing we'll do on the Mac first.
The point being that the picture is more complicated. The full email describes in some detail why Mac Office should continue to be supported, despite its low profitability at the time. The linked Macworld article hides all of that and pretends that this was an attack on Apple. It wasn't. This is why you should always try to go to the original source, not someone else's agenda based report of it.
I don't know if you intended this or not, but there's sort of a veiled implication in your post that the Win32 API lives in the kernel. It does not. In fact, the Win32 API is a pure userland API, and calls through to a much smaller set of kernel syscalls when necessary. So a Win32 API call does not equate to a 'syscall' in the normal sense, as such a call may make 0 or more calls to the kernel depending on what needs to be done.
Incidentally, this story is completely and utterly MIA from the major news organizations. Not one has even vaguely alluded to it. Just check Google News.
Good thing? Bad thing? I don't know. But a little odd, if you ask me.
Disregarding the fact that "gdi.exe" doesn't exist in the NT ancestral line, the new functions are accessible quite easily. Pick up the most recent Platform SDK, and look up the docs. There's a number of defines that activate functions only available in newer versions of Windows. They're not enabled by default so you don't inadvertently make your app not compatible with Win2k, for example.
You switch when you are discontent with what you have. Firefox users, and IE7 users, are not generally discontent. Plenty of IE6 users are. That's all there is to it.
Yes, it's true, and the poor Office team discovered that in the hardest way possible. Jensen Harris discusses it a bit here. (By the way, there's 8 articles in that series and they're all worth reading. Check the "Why the New UI" tag.
Adaptive Menus were not successful. In my opinion, they actually add complexity to the interface. [...] As a result, even for the Office 2007 applications that are still using old-style UI (such as Publisher, Project, and Visio), we've turned off Adaptive Menus by default.
Admittedly, it seems like this should've been obvious from day 1. But UI is really not a cut and dried thing; they made a mistake and they're fixing it.
I'm sure Steve was "surprised and disappointed" too. Apple was apparently talking with Cisco all that time, just to have Cisco actually ship a product with the name just a month before the MacWorld keynote
Had you read the article, you would have known that Cisco has been shipping an IPhone product since it bought InfoGear in 2000, and InfoGear was shipping it in 1996.
They have a full decade of an active product with the name before Apple's announcement. This wasn't some Cisco ambush.
The algorithm is actually stunningly simple:
if( applicant.GPA > 3.8 || applicant.GPA
Why the second condition? Well, anybody with a GPA that low must be an eccentric genius, because nobody else would have the balls to apply to Google with that.
Now for example I have had a gmail account for years, however many of my friends (non technical, and particularly the legally minded) rejected invites after reading the privacy conditions.
Well, if you find multi minute loading times acceptable, and think throwing away the ability to stream efficiently is fine, then by all means go ahead. I'll be over here, building real graphics systems and not toys.
Parent is right. Collada isn't an end-all format. It's meant as an exchange and interop format. You'd have to be an idiot to write a game that uses Collada at runtime. But for feeding between various modelers, processing tools, and the like, it's a perfect format.
Right tool for the job etc etc. You know the drill.
Vista is Cairo. As is Win2K, and Active Directory, and Exchange, and SQL Server, and god knows how many other technologies out of that company. It dissolved out of product status long ago, but the overarching goals are really what have driven Windows and related development for close to 15 years. Vista was a badly executed push towards that same ideal. That "information at your fingertips" motto has controlled Windows development. Just look at the sheer prevalence of search in Vista. (Some people will want to bring up Apple's Spotlight here. Yes, they are extremely similar. Vista's version existed internally for a long time, though. Things like this tend to come up independently in multiple companies around the same time.)
Back to the point, this article is a mess. The starting premise, though he tries to hide it, is that Vista is bad mmkay. Everything else is built backwards from there. That's why everything else is so akwardly set up and the connections are so tenuous. He's justifying his own baseless anti-Vista sentiment; the Cairo thing is just a distraction. (And let's face it, there's enough real problems with Vista that this straw man is just sad.)
This is a common instinct for a software developer -- toss up a dialog of some sort and let the user sort it out.
That doesn't work.
Raymond Chen has a lot more to say on the matter, but the basic idea is this. Most users do not give a damn what the dialog box says. It is in the way of something they are trying to do. They will simply attempt to get rid of the dialog as quickly as possible. Then when something does go wrong, they will call up support, and "your operating system isn't supported" is most likely to simply piss off the customer at this point.
MS has figured this out the painful way, by spending billions of UI R&D money, and by making the mistakes and being clobbered by the results. Learn from them.
Normally, yes. But this isn't a normal situation. Exceptions are relying on the stack (and a specific register on x86, I forget which) to do the right thing. The problem is that you destroyed the thread stack in order to get arbitrary code execution in the first place. So there's no telling where an exception will take you, because at least one stack frame is damaged.
To make matters worse -- or better, depending on who you are -- when a system exception happens, the kernel goes into "uh oh something's wrong" mode. That mode includes fairly aggressive verification of each and every stack frame. So it's very possible that the kernel will notice the damage to the stack and forcibly end the process.
The problem lies in the subtlety of "not successful()" in your psuedo-code. It can't be implemented, to be exact. What you're generally trying to do in a buffer overflow attack is to replace the return address for the current function with the address for the code you actually want to run. If you get your addresses wrong, you crash and you're done. And when you're playing games with this sort of thing, exceptions are pretty much out the window. You can't rely on using SEH (structured exception handling, look it up if you're not familiar) to save your ass, because guess what -- you destroyed the application stack to get here! If you take an exception, you're completely gone.
So basically there's no reliable way to actually execute the desired code. All of the solutions you'd normally apply, thinking from an apps programmer's point of view, no longer work. Remember, the virus is a parasite which will destroy the process beyond repair. The goal is to jump ship and set up somewhere core to the system. And none of the usual mechanisms are functional because you've gone and mucked them up. You need to talk (in)directly to the operating system, and ASLR makes that impossible to do reliably.
That's the theory, anyway. Hackers have proven to be rather clever and innovative.
The URL is of course hashed before it's transmitted anywhere. So the only sites they can actually recognize are ones that exist in their database as phishing sites. Maybe not hugely comforting, but it's not like they're blasting your browsing behavior back to their servers in plaintext.
Course that didn't stop me from turning it off anyway. I guess there are a lot of retards out there, but I'm not one.
The slashot summary is deceptive (probably deliberately). From TFA:
Do not open or save Word files that you receive from un-trusted sources or that you receive unexpectedly from trusted sources.
The point is that there is a danger that a trojan on someone else's machine could start spreading infected Word files inside a corporation, or just amongst friends. Note furthermore:
The vulnerability cannot be exploited automatically through e-mail. For an attack to be successful a user must open an attachment that is sent in an e-mail message.
Users whose accounts are configured to have fewer user rights on the system could be less affected than users who operate with administrative user rights.
It can't be triggered automatically, and limited accounts (like every Vista system) will be largely unaffected. (Because exploits will usually try to root the box or install something, both of which will be prevented.)
Also observe that Office 2007 isn't affected. Obviously MS is doing something right in the next generation of their products.
Slashdot Mirror
Here's a hint: it's not.
Eh. Think back to the blog about the start menu->shut down menu thing that took a year to develop. Communication between groups in Windows is not very good, and so when the Vista guys fix a vulnerability there, the XP guys don't necessarily get the news. It's not exactly a redeeming point, I know, but I doubt there's anything malicious or intentional going on. Like most things, it's just disorganization and red tape.
Wow, you're a fucking retard.
I agree that the drivers are at fault; this reeks of a mistake on Intel's part inside the GMA driver. All of the video card manufacturers have been going through hell working on getting things working smoothly in the new driver model. While Apple engineers are generally fairly incompetent at doing actual Windows coding, I'm firmly in the camp that an application should never be able to cause a BSOD.
Well. Let me show you a screenshot I snapped early on in the IE betas. I think this was beta 1, but it's been a while now. And just to be clear, I took the shot myself. It's not doctored in any way. It happened while I was on the MS corporate network, and you can even see that the Passport login passed. Flagged for Phishing
Very levelheaded. How would the comments read if it were Apple accusing Microsoft, I wonder.
Wait, I guess we already know the answer to that.
The point being that the picture is more complicated. The full email describes in some detail why Mac Office should continue to be supported, despite its low profitability at the time. The linked Macworld article hides all of that and pretends that this was an attack on Apple. It wasn't. This is why you should always try to go to the original source, not someone else's agenda based report of it.
I don't know if you intended this or not, but there's sort of a veiled implication in your post that the Win32 API lives in the kernel. It does not. In fact, the Win32 API is a pure userland API, and calls through to a much smaller set of kernel syscalls when necessary. So a Win32 API call does not equate to a 'syscall' in the normal sense, as such a call may make 0 or more calls to the kernel depending on what needs to be done.
Incidentally, this story is completely and utterly MIA from the major news organizations. Not one has even vaguely alluded to it. Just check Google News.
Good thing? Bad thing? I don't know. But a little odd, if you ask me.
Disregarding the fact that "gdi.exe" doesn't exist in the NT ancestral line, the new functions are accessible quite easily. Pick up the most recent Platform SDK, and look up the docs. There's a number of defines that activate functions only available in newer versions of Windows. They're not enabled by default so you don't inadvertently make your app not compatible with Win2k, for example.
You switch when you are discontent with what you have. Firefox users, and IE7 users, are not generally discontent. Plenty of IE6 users are. That's all there is to it.
They have a full decade of an active product with the name before Apple's announcement. This wasn't some Cisco ambush.
The algorithm is actually stunningly simple:
if( applicant.GPA > 3.8 || applicant.GPA Why the second condition? Well, anybody with a GPA that low must be an eccentric genius, because nobody else would have the balls to apply to Google with that.
Your friends read privacy conditions?! Seriously?
Oh wait. Do you attend law school, maybe?
Well, if you find multi minute loading times acceptable, and think throwing away the ability to stream efficiently is fine, then by all means go ahead. I'll be over here, building real graphics systems and not toys.
Parent is right. Collada isn't an end-all format. It's meant as an exchange and interop format. You'd have to be an idiot to write a game that uses Collada at runtime. But for feeding between various modelers, processing tools, and the like, it's a perfect format.
Right tool for the job etc etc. You know the drill.
Parent is right, this article is just...bitter.
Vista is Cairo. As is Win2K, and Active Directory, and Exchange, and SQL Server, and god knows how many other technologies out of that company. It dissolved out of product status long ago, but the overarching goals are really what have driven Windows and related development for close to 15 years. Vista was a badly executed push towards that same ideal. That "information at your fingertips" motto has controlled Windows development. Just look at the sheer prevalence of search in Vista. (Some people will want to bring up Apple's Spotlight here. Yes, they are extremely similar. Vista's version existed internally for a long time, though. Things like this tend to come up independently in multiple companies around the same time.)
Back to the point, this article is a mess. The starting premise, though he tries to hide it, is that Vista is bad mmkay. Everything else is built backwards from there. That's why everything else is so akwardly set up and the connections are so tenuous. He's justifying his own baseless anti-Vista sentiment; the Cairo thing is just a distraction. (And let's face it, there's enough real problems with Vista that this straw man is just sad.)
This is a common instinct for a software developer -- toss up a dialog of some sort and let the user sort it out.
That doesn't work.
Raymond Chen has a lot more to say on the matter, but the basic idea is this. Most users do not give a damn what the dialog box says. It is in the way of something they are trying to do. They will simply attempt to get rid of the dialog as quickly as possible. Then when something does go wrong, they will call up support, and "your operating system isn't supported" is most likely to simply piss off the customer at this point.
MS has figured this out the painful way, by spending billions of UI R&D money, and by making the mistakes and being clobbered by the results. Learn from them.
Normally, yes. But this isn't a normal situation. Exceptions are relying on the stack (and a specific register on x86, I forget which) to do the right thing. The problem is that you destroyed the thread stack in order to get arbitrary code execution in the first place. So there's no telling where an exception will take you, because at least one stack frame is damaged.
To make matters worse -- or better, depending on who you are -- when a system exception happens, the kernel goes into "uh oh something's wrong" mode. That mode includes fairly aggressive verification of each and every stack frame. So it's very possible that the kernel will notice the damage to the stack and forcibly end the process.
Nope, not so easy.
The problem lies in the subtlety of "not successful()" in your psuedo-code. It can't be implemented, to be exact. What you're generally trying to do in a buffer overflow attack is to replace the return address for the current function with the address for the code you actually want to run. If you get your addresses wrong, you crash and you're done. And when you're playing games with this sort of thing, exceptions are pretty much out the window. You can't rely on using SEH (structured exception handling, look it up if you're not familiar) to save your ass, because guess what -- you destroyed the application stack to get here! If you take an exception, you're completely gone.
So basically there's no reliable way to actually execute the desired code. All of the solutions you'd normally apply, thinking from an apps programmer's point of view, no longer work. Remember, the virus is a parasite which will destroy the process beyond repair. The goal is to jump ship and set up somewhere core to the system. And none of the usual mechanisms are functional because you've gone and mucked them up. You need to talk (in)directly to the operating system, and ASLR makes that impossible to do reliably.
That's the theory, anyway. Hackers have proven to be rather clever and innovative.
There are only two things I hate: Those who are intolerant of other people's cultures, and the Dutch.
The URL is of course hashed before it's transmitted anywhere. So the only sites they can actually recognize are ones that exist in their database as phishing sites. Maybe not hugely comforting, but it's not like they're blasting your browsing behavior back to their servers in plaintext.
Course that didn't stop me from turning it off anyway. I guess there are a lot of retards out there, but I'm not one.
Also observe that Office 2007 isn't affected. Obviously MS is doing something right in the next generation of their products.
Can they do it in minutes when your codebase is in excess of 50 million lines of code?