Nevermind. I've read through some of your comment history, and I've discovered that you're not the kind of person who's likely to be capable of intelligent, rational, cordial debate. You might the the first two, but you're a jerk while you're doing it. I'm not interested in hearing your responses or debating with you further.
Does it? Are you guessing, or do you know? A clean install of Vista SP0 does not have any services listening through the firewall, and an nmap scan of the host does not find anything. I'd be shocked if this was changed for SP2, but I'll know once the install is done.
Support does not mean what you are claiming
Can you point me to Microsoft's definition of support, then? I couldn't find it with a brief search. Does it mean, "If we deign to release an update, you may download it?"
http://support.microsoft.com/gp/lifepolicy isn't really useful for my question, because it uses circular definitions. "Extended support" means you get "security update support" (this falls under a security vulnerability) but they never say what they consider "support" to mean.
Also, by not patching XP, they're saying that you shouldn't have any services listening on XP. I might be able to buy that argument if the same was said for Vista, however they patched Vista. Both are OS that are used for workstations and home support (i.e., not a server OS.) Both have similar services that can run (e.g., file and print sharing).
Also, the name-calling seems unnecessary. Why can't we have civil discourse anymore? If I'm misunderstanding the definition of "support", why not enlighten me? And no:
Support doesnt mean "microsoft is my bitch"
doesn't really count. You've told me what support isn't, not what it is.
Keeping a password secret is not obscurity. Security through obscurity classically refers to wanting to keep an algorithm secret, not wanting to keep passwords secret.
Well I didn't. I don't even have trillions to blast.
But yeah. The war, a thinly veiled attack on terrorism, was pretty much exactly there to protect US interests, keep the dollar afloat, and maintain regimes that will be friendly enough to supply us with oil. It's fundamentally in line with what I said--we're protecting our own (not from being attacked, but protecting the status quo.)
I'm not making a moral judgment about the war, incidentally. The war is in line with protecting our own, but in that case, I think it's going too far.
The universal healthcare issue is really broken, but no society is perfect.
America has that same childish and ignorant "for mother country" thing going on as well
If we had international laws, policies, standards of living, etc. I'd agree with you. As we don't, I don't see a problem with wanting to take care of our own. International espionagers aren't going to share information--they only want to take it.
It's similar to the prisoner's dilemma. We'd probably all do better overall if we all worked together. China's not going to work with us, though, which means that if we just give them the technology, we're the suckers.
Most Slashdotters don't understand what security is. Security and safety are not synonymous. Obscurity may make you safer, but it does not make you more secure.
It wouldn't be reneging. But that's the point. Two operating systems from the same vendor, the same bug affects each, both are still under support, but they refuse to patch one of them. And all of the Microsofties come out of the woodwork to defend Microsoft for not wanting to support an 8 year old OS that they still sell and promised to support.
XP SP2 and later are fine by default. What does that mean? Does that mean it's the only possible configuration? Or is it reasonable that an XP SP2 computer could end up in a state where it does have a listening service configured in the client firewall? Doesn't Vista include "a stateful host firewall that provide protection for computers against incoming traffic from the Internet [...]"? I should think so, so wouldn't that invalidate their reasoning?
Exactly. They patched Vista, whose default configuration is identical, but they won't patch XP.
You might expect a dealer to service a 30 year old car, but you're probably going to have to pay through the nose for it (and I've read of at least one case where a dealer didn't have the parts to service a car because it was so old).
Not necessarily. If it's a manufacturing defect which warrants a recall, it may not matter if it's out of warranty. My 5 year old truck (out of warranty for 3 years) was recalled to fix a problem where the tailgate could fall off causing injury to those nearby.
With software, it's a different sort of injury, of course, but it could still cause people lots of problems.
And besides, they still sell XP to OEMs. My 3 month old Netbook came with it. And they're saying that they won't fix this bug. That's absurd.
Similar to the reason that we don't see many new power plants these days. Old plants were grandfathered in after safety and environmental laws were enacted. New plants are held to much higher standards.
So we just have to get by on the old and busted plants which spew tons of junk in the air.
You're right--that's really the correct solution. But the submitter presumably wanted something that isn't going to require a lot of coding.
From what I understand of the VFS, this would require an extra layer under the syscalls to fake the owner. It may be possible with device mapper, but I'm not sure.
With Fat32, iirc, all files essentially have rwxrwxrwx permissions. When you double-click a file like this in Ubuntu, it doesn't know whether you want to run the file or execute it, so it has to ask you.
With ext2/3, permissions will be maintained, and thus text files won't have a +x. Ubuntu can safely assume that you don't want to execute it, so it will default to display it. However UIDs may not map perfectly from one machine to another, so if the user sancho has UID 500 on my personal machine, but when I log into my friend's machine, I'm using UID 600, then I may have restricted permissions on the files. If I don't have superuser access, I may not be able to correct the problem.
The obvious solution is to use rw-rw-rw- for files on the USB drive, however copying them to a new machine will preserve those permissions (which probably isn't what you want if you put them onto a multi-user machine.)
1) Unix files created on a Unix system and moved between Unix volumes. In this case, there's no issue. Unix doesn't track creation time, so that particular metadata isn't lost during the move. The ctime represents the inode change time, so it's correct to change it when you change the destination inode.
2) Windows files created on a Windows system and moved between Unix volumes. This was actually a point of contention in the past. POSIX doesn't have a notion of a creation time, so when writing drivers or compatibility layers (e.g. Cygwin), there's no clear way to deal with creation time attributes (that the syscalls don't have any idea how to deal with--or even that they exist.) It sounds like once you put the file onto a Unix filesystem, you lose that metadata. It's not unlike transferring a Mac OS file with resource forks to Windows or Unix. I assume something similar happens with NTFS alternate streams.
1) Ubiquitous surveillance is different from occasional, one-off surveillance. Prior to the technology boom, surveillance was hard. It essentially required manpower, and thus there was little potential for abuse. A police officer with a vendetta against someone must physically watch them all of the time, meaning that they're either spending vast amounts of personal time or are not doing their job (or both.) With video cameras, the manpower requirement is significantly reduced. A police officer with a vendetta against someone, for example, could easily stalk them using the cameras without having to follow them constantly.
2) The balance of power is vastly different when it's the government doing the surveillance against a citizen compared to a citizen doing the surveillance against another citizen.
3) I should be able to record my day-to-day activities. Surveillance cameras do not record the day-to-day activities of the operator.
Slashdot Mirror
Nevermind. I've read through some of your comment history, and I've discovered that you're not the kind of person who's likely to be capable of intelligent, rational, cordial debate. You might the the first two, but you're a jerk while you're doing it. I'm not interested in hearing your responses or debating with you further.
Have a nice day.
but a freshly installed Vista + SP2 does
Does it? Are you guessing, or do you know? A clean install of Vista SP0 does not have any services listening through the firewall, and an nmap scan of the host does not find anything. I'd be shocked if this was changed for SP2, but I'll know once the install is done.
Support does not mean what you are claiming
Can you point me to Microsoft's definition of support, then? I couldn't find it with a brief search. Does it mean, "If we deign to release an update, you may download it?"
http://support.microsoft.com/gp/lifepolicy isn't really useful for my question, because it uses circular definitions. "Extended support" means you get "security update support" (this falls under a security vulnerability) but they never say what they consider "support" to mean.
Also, by not patching XP, they're saying that you shouldn't have any services listening on XP. I might be able to buy that argument if the same was said for Vista, however they patched Vista. Both are OS that are used for workstations and home support (i.e., not a server OS.) Both have similar services that can run (e.g., file and print sharing).
Also, the name-calling seems unnecessary. Why can't we have civil discourse anymore? If I'm misunderstanding the definition of "support", why not enlighten me? And no:
Support doesnt mean "microsoft is my bitch"
doesn't really count. You've told me what support isn't, not what it is.
Keeping a password secret is not obscurity. Security through obscurity classically refers to wanting to keep an algorithm secret, not wanting to keep passwords secret.
I would argue that the door is equally secure in both places, but that your key management system needs some serious work in the first situation.
The area behind the door is not secure in the first scenario using the definition of secure to mean "free from danger or harm."
I see what you did there.
Well I didn't. I don't even have trillions to blast.
But yeah. The war, a thinly veiled attack on terrorism, was pretty much exactly there to protect US interests, keep the dollar afloat, and maintain regimes that will be friendly enough to supply us with oil. It's fundamentally in line with what I said--we're protecting our own (not from being attacked, but protecting the status quo.)
I'm not making a moral judgment about the war, incidentally. The war is in line with protecting our own, but in that case, I think it's going too far.
The universal healthcare issue is really broken, but no society is perfect.
America has that same childish and ignorant "for mother country" thing going on as well
If we had international laws, policies, standards of living, etc. I'd agree with you. As we don't, I don't see a problem with wanting to take care of our own. International espionagers aren't going to share information--they only want to take it.
It's similar to the prisoner's dilemma. We'd probably all do better overall if we all worked together. China's not going to work with us, though, which means that if we just give them the technology, we're the suckers.
To be fair, when debating, it's up to the person putting forth the argument to support it.
Most Slashdotters don't understand what security is. Security and safety are not synonymous. Obscurity may make you safer, but it does not make you more secure.
Where's the distinction? They chose to patch the same bug in another OS, so they must think it's worthy of being patched.
Selectively patching critical bugs is not support.
It wouldn't be reneging. But that's the point. Two operating systems from the same vendor, the same bug affects each, both are still under support, but they refuse to patch one of them. And all of the Microsofties come out of the woodwork to defend Microsoft for not wanting to support an 8 year old OS that they still sell and promised to support.
If you drop your subscription, you lose the rights to the software though. Also, they're not for general purpose use--they're for testing.
Microsoft has support for XP through 2014. They're reneging on that by refusing to patch this flaw.
XP SP2 and later are fine by default. What does that mean? Does that mean it's the only possible configuration? Or is it reasonable that an XP SP2 computer could end up in a state where it does have a listening service configured in the client firewall? Doesn't Vista include "a stateful host firewall that provide protection for computers against incoming traffic from the Internet [...]"? I should think so, so wouldn't that invalidate their reasoning?
Exactly. They patched Vista, whose default configuration is identical, but they won't patch XP.
You might expect a dealer to service a 30 year old car, but you're probably going to have to pay through the nose for it (and I've read of at least one case where a dealer didn't have the parts to service a car because it was so old).
Not necessarily. If it's a manufacturing defect which warrants a recall, it may not matter if it's out of warranty. My 5 year old truck (out of warranty for 3 years) was recalled to fix a problem where the tailgate could fall off causing injury to those nearby.
With software, it's a different sort of injury, of course, but it could still cause people lots of problems.
And besides, they still sell XP to OEMs. My 3 month old Netbook came with it. And they're saying that they won't fix this bug. That's absurd.
Both Vista and Windows 7 were sold as pre-orders for a reduced cost. In fact, Windows 7 is doing better than Vista at pre-orders:
http://www.crunchgear.com/2009/07/15/in-8-hours-windows-7-pre-orders-overtake-vista-pre-orders/
Similar to the reason that we don't see many new power plants these days. Old plants were grandfathered in after safety and environmental laws were enacted. New plants are held to much higher standards.
So we just have to get by on the old and busted plants which spew tons of junk in the air.
Fantastic idea! I do try to be a kinder, gentler grammar/spelling nazi.
FYI:
http://en.wikipedia.org/wiki/Per_se
You're right--that's really the correct solution. But the submitter presumably wanted something that isn't going to require a lot of coding.
From what I understand of the VFS, this would require an extra layer under the syscalls to fake the owner. It may be possible with device mapper, but I'm not sure.
The question is poorly worded/explained.
With Fat32, iirc, all files essentially have rwxrwxrwx permissions. When you double-click a file like this in Ubuntu, it doesn't know whether you want to run the file or execute it, so it has to ask you.
With ext2/3, permissions will be maintained, and thus text files won't have a +x. Ubuntu can safely assume that you don't want to execute it, so it will default to display it. However UIDs may not map perfectly from one machine to another, so if the user sancho has UID 500 on my personal machine, but when I log into my friend's machine, I'm using UID 600, then I may have restricted permissions on the files. If I don't have superuser access, I may not be able to correct the problem.
The obvious solution is to use rw-rw-rw- for files on the USB drive, however copying them to a new machine will preserve those permissions (which probably isn't what you want if you put them onto a multi-user machine.)
Ah, so there are a couple of scenarios here.
1) Unix files created on a Unix system and moved between Unix volumes. In this case, there's no issue. Unix doesn't track creation time, so that particular metadata isn't lost during the move. The ctime represents the inode change time, so it's correct to change it when you change the destination inode.
2) Windows files created on a Windows system and moved between Unix volumes. This was actually a point of contention in the past. POSIX doesn't have a notion of a creation time, so when writing drivers or compatibility layers (e.g. Cygwin), there's no clear way to deal with creation time attributes (that the syscalls don't have any idea how to deal with--or even that they exist.) It sounds like once you put the file onto a Unix filesystem, you lose that metadata. It's not unlike transferring a Mac OS file with resource forks to Windows or Unix. I assume something similar happens with NTFS alternate streams.
There is no 'creation time' attribute in Linux. Are you perhaps looking at 'ctime' and thinking that stands for 'create'?
stat filename will show you that it actually stands for 'change'.
There are several issues here.
1) Ubiquitous surveillance is different from occasional, one-off surveillance. Prior to the technology boom, surveillance was hard. It essentially required manpower, and thus there was little potential for abuse. A police officer with a vendetta against someone must physically watch them all of the time, meaning that they're either spending vast amounts of personal time or are not doing their job (or both.) With video cameras, the manpower requirement is significantly reduced. A police officer with a vendetta against someone, for example, could easily stalk them using the cameras without having to follow them constantly.
2) The balance of power is vastly different when it's the government doing the surveillance against a citizen compared to a citizen doing the surveillance against another citizen.
3) I should be able to record my day-to-day activities. Surveillance cameras do not record the day-to-day activities of the operator.
Moreover, it's illegal to cover your face for the purpose of disguising or hiding your identity in many places.