The question becomes, now that this technology is cheaper, will my VP be buying new Windows Mobile enabled cell phones for his entire department just so we can put in more hours?
That's a pretty stupid question. Phrased that way, the VP would be an idiot unless he has a chunk of change for over time or wants to be sure that he doesn't have some stupid IT people show up at a company function (muwahaha, I'll have them install stuff during the party so they can't show up - muwaaahhhaaahaha).
What the real questions are these: Is the value proposition greater or less than what we have? If it is greater value, then how long to pay out? Longer than the mean time to obsolescence of the new devices? What's the NPV? Should we move to an integrated solution, or stay on the best of breed? How many of the Blackberry devices are out there now in our company and what is the growth rate of the demand for them?
Oh wait - those questions aren't inflamatory so the article wouldn't have gotten posted if they were asked instead of the silly one...
Scientists have tested this one over and over. It has been called many things, but the bottom line is that if people co-operate (give their $1 in your idea) they can get something. But if they are greedy, they have a chance for much more. It always works out that people DON'T co-operate. Why pay my $1 when I can hope that a bunch of "suckers" will pay and I get it free?
Funny, I thought Windows had worked on multi-proc systems for years. I really doubt that a bunch of serious new issues will crop up just because they stuck two chips on one die. I mean they still use the same slow memory interconnect (OK, out of my league here - but it's something like northbridge or southbridge or londonbridge). Anyway, this isn't as much of a new architecture as people make it out to be...
First, the article isn't hidden and available to just MS and OEM's. I can see it fine through Premier Online (requires a select or enterprise agreement granted). The article also says NOTHING about any processor architecture - it is ALL of them - AMD, Intel, dual core, single core, hyperthreading - doesn't matter. According to the article, it is anything runnning XPSP2.
I agree that the machines are flawed - however a minor correction. The machines don't threaten anything. It's the criminals who attempt to hack them who threaten the democratic process. The same criminals that used to see to it that dead people voted every year now just have an easier and more sure way to rig the election. That's not the machine threatening anything. If people were honest, these machines would be fine. Since people are not honest, people threaten the system.
While government in general is massively inefficient, the true "cost" of your proposal to "rescind" everything (even were it legal) would actually be - no police, no firemen, roads getting crappy even quicker, massive theft and robberies since social programs wouldn't be funded, etc.
Seems to me an artificial singularity (read black hole) could be sent on a couple of different orbit tracks and grab stuff. This would work for cases that magnets wouldn't (for instance large paint flakes that are still dangerous at umpteen hundred MPH / KPH, aluminum, etc.)
Now we just need to make contact with the Romulans - I seem to recall that their warp cores used just the type of singularity we would need...
It isn't TOO bad because of the built in file and registry virtualization in Vista. If a program running with a LUA token tries to write to say the "C:\Program Files\PoorlyWrittenApp" folder, that write will result in a copy of the file (if it already existed) being made and placed in a location under the user's profile. Then the write to that file will succeed in the new location in the user profile. The OS will preferentially read that new file whenever the file in program files is being "read" by the app.
The same thing works for registry entries. There are certain files (like.exe, etc.) that are never virtualized to make sure people don't get DoS attacked by "replacing" their exe files. There are API's for application developers to specify that they don't want certain files, folders, or registry keys to virtualize. All in all, it makes the app compat story pretty robust.
No, that's not right. For instance Office 2003 runs fine as a LUA user. There are many apps that do, it just isn't enough yet. There are some cool technologies in Vista to allow more apps to work as LUA though. File and Registry virtualization will allow writes to protected locations to succeed - but virtualize it to a place under the user profile and then preferentially read from there. Obviously this comes with its own set of problems but overall will help to make apps that still are badly behaved able to run as LUA. So far, the default user account in Vista is a "Protected Admin" which would be a user with a LUA token who can get a popup box asking them to "elevate" when performing an admin task. If they choose to elevate, the process requiring the extra rights will use a token with admin rights. I'm still hoping that they go to full LUA (person who can only elevate with another account and password), because people may start clicking "Yes, elevate" all the time. The situation looks to be much, much better than it was with older versions of Windows though.
OK, who the heck mod'ed the parent insightful? Sure, let's break everything! That'll work! While it might be a nice idea; the type of thing you throw around in a brain-storm session where "nothing is stupid", there is no way it is even close to possible to remove it now.
The best that could be done is to change the behavior a bit each rev (Vista starts this by the way) to make it very hard to install ActiveX and eventually very hard to run it. Maybe in the OS after Blackcomb they can finally get rid of it.
No, that isn't the case. Again, you are finding the user mode rootkits that way. They are only hiding from ntdll.dll (and hence Explorer.exe doesn't show them, cmd.exe doesn't show them). The redirector is running as system, so the user mode ones can't hide from that. This is why you can see them over remote mounted disks (C$,etc.).
However, if you read up on the kernel mode ones (some of the talks Mark Russinovich has given -like at Tech Ed this year), you'll see that these touch the kernel itself and the redirector will not expose them (so C$, etc. won't work).
It's just a matter of different architectures and different methods of "rooting" a machine.
Sorry, but that isn't 100% correct. The USER MODE rootkits can be found in the manner you mentioned - mapping a drive to the C$ from another machine.
However, the redirector still has to use the Kernel so a kernel mode rootkit will NOT be found by your method. In fact, the SONY rookit will not be found by that method.
To find kernel mode rootkits, you need full offline scanning - booting to WinPE or Knoppix or something. If the OS is booted and you are just remoting it, the kernel mode one will still lie to you.
Right, that's true. I find that on my desktop machine, having it enabled allows me to more seamlessly run VMWare virtual machines. One can be spinning one virtual processor fairly well and not freeze my host machine's apps. Of course this only goes so far; you do still feel it - but you aren't totally processor starved. So far on my development desktops it has been a good thing to have HT.
Maybe you missed out, but rootkits were something developed for Unix first. They've been around a long time on both Unix and Windows and have always been difficult to detect. Maybe we should be demanding a fix - but it would be from just about all the OS makers out there.
Is the lack of polish you see just due to them trying to make it look very very similar on all platforms (kind of the least common denominator thing so that they can't use the coolest KDE stuff, the coolest Gnome stuff, or the coolest Windows stuff)? Or was it something else? For me it was the speed (or lack thereof) and the fact that my time is worth more than MS Office costs (I have a lot of macros). I have OO2 on my SuSe box, but I'm not about to run it on my Windows box.
You are right of course. The only thing I would add is that I HATE it when the ads take up most of the page and I only get a single column of text about 35 characters wide that I have to scroll forever to read.
Slashdot Mirror
The question becomes, now that this technology is cheaper, will my VP be buying new Windows Mobile enabled cell phones for his entire department just so we can put in more hours?
That's a pretty stupid question. Phrased that way, the VP would be an idiot unless he has a chunk of change for over time or wants to be sure that he doesn't have some stupid IT people show up at a company function (muwahaha, I'll have them install stuff during the party so they can't show up - muwaaahhhaaahaha).
What the real questions are these: Is the value proposition greater or less than what we have? If it is greater value, then how long to pay out? Longer than the mean time to obsolescence of the new devices? What's the NPV? Should we move to an integrated solution, or stay on the best of breed? How many of the Blackberry devices are out there now in our company and what is the growth rate of the demand for them?
Oh wait - those questions aren't inflamatory so the article wouldn't have gotten posted if they were asked instead of the silly one...
Scientists have tested this one over and over. It has been called many things, but the bottom line is that if people co-operate (give their $1 in your idea) they can get something. But if they are greedy, they have a chance for much more. It always works out that people DON'T co-operate. Why pay my $1 when I can hope that a bunch of "suckers" will pay and I get it free?
Funny, I thought Windows had worked on multi-proc systems for years. I really doubt that a bunch of serious new issues will crop up just because they stuck two chips on one die. I mean they still use the same slow memory interconnect (OK, out of my league here - but it's something like northbridge or southbridge or londonbridge). Anyway, this isn't as much of a new architecture as people make it out to be...
First, the article isn't hidden and available to just MS and OEM's. I can see it fine through Premier Online (requires a select or enterprise agreement granted). The article also says NOTHING about any processor architecture - it is ALL of them - AMD, Intel, dual core, single core, hyperthreading - doesn't matter. According to the article, it is anything runnning XPSP2.
I agree that the machines are flawed - however a minor correction. The machines don't threaten anything. It's the criminals who attempt to hack them who threaten the democratic process. The same criminals that used to see to it that dead people voted every year now just have an easier and more sure way to rig the election. That's not the machine threatening anything. If people were honest, these machines would be fine. Since people are not honest, people threaten the system.
While government in general is massively inefficient, the true "cost" of your proposal to "rescind" everything (even were it legal) would actually be - no police, no firemen, roads getting crappy even quicker, massive theft and robberies since social programs wouldn't be funded, etc.
Seems to me an artificial singularity (read black hole) could be sent on a couple of different orbit tracks and grab stuff. This would work for cases that magnets wouldn't (for instance large paint flakes that are still dangerous at umpteen hundred MPH / KPH, aluminum, etc.)
Now we just need to make contact with the Romulans - I seem to recall that their warp cores used just the type of singularity we would need...
It isn't TOO bad because of the built in file and registry virtualization in Vista. If a program running with a LUA token tries to write to say the "C:\Program Files\PoorlyWrittenApp" folder, that write will result in a copy of the file (if it already existed) being made and placed in a location under the user's profile. Then the write to that file will succeed in the new location in the user profile. The OS will preferentially read that new file whenever the file in program files is being "read" by the app.
.exe, etc.) that are never virtualized to make sure people don't get DoS attacked by "replacing" their exe files. There are API's for application developers to specify that they don't want certain files, folders, or registry keys to virtualize. All in all, it makes the app compat story pretty robust.
The same thing works for registry entries.
There are certain files (like
No, that's not right. For instance Office 2003 runs fine as a LUA user. There are many apps that do, it just isn't enough yet. There are some cool technologies in Vista to allow more apps to work as LUA though. File and Registry virtualization will allow writes to protected locations to succeed - but virtualize it to a place under the user profile and then preferentially read from there. Obviously this comes with its own set of problems but overall will help to make apps that still are badly behaved able to run as LUA. So far, the default user account in Vista is a "Protected Admin" which would be a user with a LUA token who can get a popup box asking them to "elevate" when performing an admin task. If they choose to elevate, the process requiring the extra rights will use a token with admin rights. I'm still hoping that they go to full LUA (person who can only elevate with another account and password), because people may start clicking "Yes, elevate" all the time. The situation looks to be much, much better than it was with older versions of Windows though.
OK, who the heck mod'ed the parent insightful? Sure, let's break everything! That'll work! While it might be a nice idea; the type of thing you throw around in a brain-storm session where "nothing is stupid", there is no way it is even close to possible to remove it now.
The best that could be done is to change the behavior a bit each rev (Vista starts this by the way) to make it very hard to install ActiveX and eventually very hard to run it. Maybe in the OS after Blackcomb they can finally get rid of it.
Can you just imagine how long they stared at the CONTAINER zen-like before opening it? Must have been a LOT longer than 10 days, huh?
Obviously the moondust is detrimental to the IQ as well as the lungs.
No, that isn't the case. Again, you are finding the user mode rootkits that way. They are only hiding from ntdll.dll (and hence Explorer.exe doesn't show them, cmd.exe doesn't show them). The redirector is running as system, so the user mode ones can't hide from that. This is why you can see them over remote mounted disks (C$,etc.).
However, if you read up on the kernel mode ones (some of the talks Mark Russinovich has given -like at Tech Ed this year), you'll see that these touch the kernel itself and the redirector will not expose them (so C$, etc. won't work).
It's just a matter of different architectures and different methods of "rooting" a machine.
Sorry, but that isn't 100% correct. The USER MODE rootkits can be found in the manner you mentioned - mapping a drive to the C$ from another machine.
However, the redirector still has to use the Kernel so a kernel mode rootkit will NOT be found by your method. In fact, the SONY rookit will not be found by that method.
To find kernel mode rootkits, you need full offline scanning - booting to WinPE or Knoppix or something. If the OS is booted and you are just remoting it, the kernel mode one will still lie to you.
Yes, but "Are you of the body?"
Right, that's true. I find that on my desktop machine, having it enabled allows me to more seamlessly run VMWare virtual machines. One can be spinning one virtual processor fairly well and not freeze my host machine's apps. Of course this only goes so far; you do still feel it - but you aren't totally processor starved. So far on my development desktops it has been a good thing to have HT.
I'm lauging at the superior intellect...
You sir, will soon be sued for 2 Million dollars too. That is, if that shady company of which you speak reads /.
Maybe you missed out, but rootkits were something developed for Unix first. They've been around a long time on both Unix and Windows and have always been difficult to detect. Maybe we should be demanding a fix - but it would be from just about all the OS makers out there.
You're right: Cubicles aren't evil. It's the stockholders and upper management who try to put us in cubicles that are evil...
Is the lack of polish you see just due to them trying to make it look very very similar on all platforms (kind of the least common denominator thing so that they can't use the coolest KDE stuff, the coolest Gnome stuff, or the coolest Windows stuff)? Or was it something else? For me it was the speed (or lack thereof) and the fact that my time is worth more than MS Office costs (I have a lot of macros). I have OO2 on my SuSe box, but I'm not about to run it on my Windows box.
Ouch! Too right, but Stallman may appear and smite you...
But doesn't SCO own the GPL? I thought they claimed that people stole it from them? Darl McFries sued IBM over it didn't he?
If you could just organize it enough to prove it, right?
You are right of course. The only thing I would add is that I HATE it when the ads take up most of the page and I only get a single column of text about 35 characters wide that I have to scroll forever to read.