How about a true multi-user design built to be used on a hostile network.
NT has had that since day 1. Try again.
You know, something done in the Unix for a couple of decades before the FIRST MS Windows, and STILL not done right by MS after a good 1.5 decades of development on their flagship product.
If anything, NT's design is *more* "multiuser" than unix's "you're either root, or you're not" design.
Microsoft's biggest problem is that they're coming to realize that their operating system just plain was not designed with some of today's realities in mind.
One thing that hit me when I upgraded the OS on a low end iMac a few years back is that the newer release of OSX actually improved the performance dramaticly.
Largely because the performance of early OS X releases was so dismal, it's didn't have anywhere else to go but up (and it's still far from "fast").
I can't help comparing that to MS Windows which slows down with each release - probably one reason why people are still installing win98SE.
Many aspects of XP are faster than 2000. Windows 2003 is faster than both XP and 2000.
So, whereas when Apple releases yet another yearly release, I'm excited to try it and see all the nifty little gadgets they've put in there this year, when Microsoft waits three, four, even five YEARS to release another version of Windows, I'm thinking I'd better be blown-away. This rarely happens. In fact, all of the features that would have blown me away (*cough*WinFS*cough*) are steadily removed from the shipping OS every time the release date slips.
The problem with your comparison is that Windows is a mature product, whereas OS X is not. OS X releases have been relatively frequent simply because so much stuff has needed significant improvements (most notably, performance, but also things like UI). You have probably noticed OS X releases are becoming less frequent - this is happening because the product is maturing.
Comparing commercial release schedules to OSS is largely nonsensical, because the latter has none of the pressures and/or responsibilities of the former.
Well, patch Solitaire in Windows, you have to reboot (okay, slight exaggeration), leading to downtime ranging from minutes to hours (in the case of extremely large databases)
If rebooting a machine causes you problematic service downtime, your environment has fundamental problems that need to be addressed.
Patch anything but the kernel(and modules) in Linux? Just keep chugging along, perhaps restarting a single process or two, and a fairly transparent experience from the user perspective.
The difference between restarting some network service that everyone uses, and restarting an entire machine, is usually a matter of semantics.
Also note: downtime due to patches, maintenance, etc., is not counted as "downtime" as defined by Microsoft - just the rest of the world. So when you read downtime/uptime comparisons from Microsoft, ignore them. They redefine the terms.
The Real World is interested in *service* uptimes, not *server* uptimes. Scheduled maintenance, patching, etc of servers - assuming your environment is properly designed - should not have any impact on *service* availability.
Comparing individual server uptimes is the geek equivalent of comparing business card designs.
Do you really believe that Vista, something that realistically amounts to security fixes, a new and more annoying UI, and a few toolkits that exist elsewhere, [...]
If you really think that, then you need to do some more research. *Real* research, as well, not reading press releases. Vista has had a *lot* of work done under the hood.
Anyway, you need to think through things more, and look at past performance. You can't trust anything that MS says until you see it yourself. Every "revolutionary" technology that was so heavily pushed by MS propoganda has been dropped eventually.
Not really all that odd. I believe it's called a pre-sale. People do this on eBay all the time, selling items they don't yet have, but will send along when they get them.
The reason it's not odd is because the "security tweaks" are almost certainly going to be changes in the default configurations of things like user permissions, firewalling, workarounds for specific pieces of software, etc. Businesses are (or should be, at any rate) going to change these default settings to suit their own policies and environments.
...on the upside, at least Kim Beazley is more or less unelectable.
On the downside, this is something the Libs would *love* to do (and have proposed in the past). With not even token opposition from Labour, it's even more likely to get done.
Is that this now opens the door for the Libs to propose the same thing (an idea they've floated before) *and* get it implemented with nary a peep of dissention from politicians.
Yes, the web as well. Neither HTTP caches nor indexing search engines are recent developments. They've been around basically as long as the web has been commercialised and *certainly* as long as it has been popular.
One could conceivably - just - make the argument that people who posted content to the web in its infancy did not realise that their content was going to be publically available and copied by others for the purposes of searching and/or archiving. It certainly wouldn't apply to anything published in (at least) the last ten years, however.
OK, show me ONE legal system that says copyright is terminated and no longer exist as soon as you publish something on the net! Just one is OK!!
Firstly, I never made any such claim. I merely said that publishing content on the public internet is done so with - at the very least - knowledge that it will be copied, indexed and made publically available without restriction and - more likely - published with the *intent* of that happening.
Secondly, it seems the ruling in this case has pretty much said what you're complaining about is legal (or, at worst, not illegal) - and a good thing, too.
Just because anyone can come and see your page or what you present on a web page (for as long as it is there only though), does not mean anyone in turn are allowed to start copying and distributing the content.
Well if they couldn't, both HTTP caches and indexing search engines (amongst numerous other fundamental aspects of the internet) would not only be illegal, but impossible. AFAIK, no court has ruled in this fashion.
There is no copyright law saying so either, feel free to check all of them.
You keep flicking back and forth between asking me what I think and asking me what the law says. Please make up your mind which you want to discuss.
Someone that show something on the net does NOT tell everyone to copy and distribute it in any way, well there are probably some pages that specifically states so but not in general.
Unfortunately, the "Windows Mentality" deems that you don't feedback to software creators - instead, you just hand over money & have a shrink-wrapped box put in your hands. And when you try out the software, it either does what you want it to do or doesn't; if it's the latter, you just use it, put up with it & wait for the next version...
When was the last time you sent a developer feedback about their Windows software ?
Very few people actually use windows- you ask them what kind of computer they have, and you'll hear "Dell" or "Packard Bell" or "Gateway" - maybe even an "IBM".
By this (ridiculous) logic, "very few people actually use" cars.
So how was McDonalds responsible when the woman spilled the hot coffee in her lap, and M got sued for $1.0M. So now McDonalds have to put hot warnings on their coffee. Is is not reasonable to assume that unless someone is really "stupid", they would know that coffee is served hot.
This is OT, however, the reason McDonalds was found liable was because they were a) serving coffee at a far higher temperature than anyone would reasonably call "hot enough" and b) because they had received numerous complaints about the excessively high temperature their coffee was being served at - and other injuries it had caused - and done nothing about it.
The woman who has burned certainly shares some of the blame for clumsily opening the coffee in a rather inadvisable fashion, but what would have been a relatively minor burn and reminder that it was a silly thing to do, instead became a very serious and debilitating injury because the McDonald's coffee in question was so much hotter than expected.
There was clearly a valid case against McDonalds, in that instance. The problem was not that the coffee was hot, the problem that it was *unnecessarily* and *unexpectedly* hot, that McDonalds knew this and that they did nothing the remedy the situation.
I agree with you to a point. Sure, the user needs to be more informed on basic security practices and nothing can defeat a secure system better than a 'stupid user'. But, every product made to interact with human beings take this as a given: "Shit happens". In cars [...]
Here's the problem: computers aren't cars. Now, while the good old "car analogy" has some valid uses, this is not one of them. A car is a device manufactured, designed and optimised to do a fairly small set of well-defined tasks. Take the car outside of its "normal" parameters (eg: overloading, driving on ice with regular tyres, using the wrong type of fuel) and the car - and possibly the user - will be damaged. This is not a failure in the engineering of the car, it is a user error.
Computers are designed to handle, essentially, completely arbitrary, end-user defined tasks. Even when you throw an operating system into the mix, which reduces the set of possible tasks slightly, you *still* have a list of potential uses which is, for all intents and purposes, infinite.
Trying to design a "computer" (and by this I'm talking about the hardware+OS combination) so that it can only perform "good" tasks and not allow anything "bad" to happen is, essentially, impossible. As long as the design requirement of "arbitrary code execution" remains, the onus is always on the end user to decide which code is "good" and which code is "bad".
The same is and should be done in computers. User accounts are a must. Patching vulnerabilities should also not be mandatory. Even for the pirates.
Problem is most malware a) doesn't really require anything more than a standard user account and b) isn't exploiting software vulnerabilities (let alone unpatched ones). Which is not to say doing this things is a worthless endeavour, merely that it's nothing close to a "fix".
Microsoft has the power to enforce better security in windows by using better default configurations (your average user won't bother hardening his OS).
This is something they can - and are - doing with Vista. But they can't do it retroactively.
The other thing we need is diversity. Having one OS to rule them all is a vulnerability within itself, it just amplifies the damage.
Build it and they will come.
Security is a process that will never practically end, and blaming the user for "Almost all" malware infestations is to much ostrich-reasoning for my likings, not that it isn't true as things are today, especially on the windows side...
It's true on all platforms and it's not going to change as long as computers are able to run arbitrary code.
should the doctor have lowered the price for this instance?
Yes. More accurately, the Doctor should not be charging so much in the first place.
did the husband break the law by stealing the medicie but leaving enough for base cost plus a little extra?
Yes. Note that this does not mean he did anything wrong.
You appear to be conflating "wrong" with "illegal" (your first question uses the term "wrong", your last the term "illegal"). "Illegal" does not imply "wrong" any more than "wrong" implies "illegal".
AH! but in the auto industry "stupid" results in a class action suits against the manufacturers. Pinto's are not supposed to explode when you drive into the back of them, but "stupid" is not supposed to drive into the back of cars in the first place.
Somehow I doubt the silly woman who drove into the back of me a few weeks ago because she was busy talking on the phone and doing her nails will have much luck in court trying to blame Mazda for it.
The vast bulk of malware infestations are like the innatentive driver causing an accident, not the person being driven into whose car explodes.
Doesn't mean there's not recalls when it turns out an accident or so is caused by defective parts. Why isn't Microsoft required to fix the holes that don't need the user?
Last I checked Microsoft *do* patch their software.
And if a particualr product is found to make it easier user-caused accidents (a la the tires on Ford Explorers), the company recalls them. Why is Windows lack of security exempt from this?
Because 99% of the time it's not lack of security in the OS, it's lack of common sense in the operator.
Of course, if you want to go down the path of "software liability", you can say goodbye to any meaningful amount of OSS acceptance.
On a more serious note, you're right: Micrsoft has no business making anti-spyware workaround programs, when they ought to be fixing their crap software (i.e., the underlying problem that allows spyware to exist in the first place) instead!
Firstly, the "underlying problem" is almost always the user.
Secondly, anti-spyware and anti-virus software is protecting against things OS-level security *can't*.
Thirdly, as long as users are able to run arbitrary software, malware will always be a problem.
Slashdot Mirror
And just to highlight how silly comparing their release cycles is, that same Friday is about the 13th anniversary of Windows NT.
If Apple are still pumping out non-trivial OS X updates every 12 months in 2014, *then* I'll be impressed.
So downtime is ok as long as it's "short enough" ?
I'm glad I'm not one of your customers. At which point does a "short" downtime become actual downtime ?
NT has had that since day 1. Try again.
You know, something done in the Unix for a couple of decades before the FIRST MS Windows, and STILL not done right by MS after a good 1.5 decades of development on their flagship product.
If anything, NT's design is *more* "multiuser" than unix's "you're either root, or you're not" design.
That's because sudo is basically just a hack to work around two fundamental unix design flaws:
1. The concept of an inherently unrestricted user; and
2. That you need to be root to do a lot of useful things.
The biggest security problem with sudo is that it doesn't *grant* you $USER's (typically root's) privileges, *it turns you into that user*.
Probably never. It's a rather large hole from a security/auditing perspective.
For example ?
Largely because the performance of early OS X releases was so dismal, it's didn't have anywhere else to go but up (and it's still far from "fast").
I can't help comparing that to MS Windows which slows down with each release - probably one reason why people are still installing win98SE.
Many aspects of XP are faster than 2000. Windows 2003 is faster than both XP and 2000.
The problem with your comparison is that Windows is a mature product, whereas OS X is not. OS X releases have been relatively frequent simply because so much stuff has needed significant improvements (most notably, performance, but also things like UI). You have probably noticed OS X releases are becoming less frequent - this is happening because the product is maturing.
Comparing commercial release schedules to OSS is largely nonsensical, because the latter has none of the pressures and/or responsibilities of the former.
If rebooting a machine causes you problematic service downtime, your environment has fundamental problems that need to be addressed.
Patch anything but the kernel(and modules) in Linux? Just keep chugging along, perhaps restarting a single process or two, and a fairly transparent experience from the user perspective.
The difference between restarting some network service that everyone uses, and restarting an entire machine, is usually a matter of semantics.
Also note: downtime due to patches, maintenance, etc., is not counted as "downtime" as defined by Microsoft - just the rest of the world. So when you read downtime/uptime comparisons from Microsoft, ignore them. They redefine the terms.
The Real World is interested in *service* uptimes, not *server* uptimes. Scheduled maintenance, patching, etc of servers - assuming your environment is properly designed - should not have any impact on *service* availability.
Comparing individual server uptimes is the geek equivalent of comparing business card designs.
If you really think that, then you need to do some more research. *Real* research, as well, not reading press releases. Vista has had a *lot* of work done under the hood.
Anyway, you need to think through things more, and look at past performance. You can't trust anything that MS says until you see it yourself. Every "revolutionary" technology that was so heavily pushed by MS propoganda has been dropped eventually.
Uh, is there any vendor this *doesn't* apply to ?
The reason it's not odd is because the "security tweaks" are almost certainly going to be changes in the default configurations of things like user permissions, firewalling, workarounds for specific pieces of software, etc. Businesses are (or should be, at any rate) going to change these default settings to suit their own policies and environments.
Ironically, the dirtiest magazines are only (legally) available for sale in Australia's capital city.
I always thought you guys were a bit more liberal and easy-going than what the news makes you out to be..
*We* are. Idiotic politicians on the other hand...
On the downside, this is something the Libs would *love* to do (and have proposed in the past). With not even token opposition from Labour, it's even more likely to get done.
Is that this now opens the door for the Libs to propose the same thing (an idea they've floated before) *and* get it implemented with nary a peep of dissention from politicians.
Yes, the web as well. Neither HTTP caches nor indexing search engines are recent developments. They've been around basically as long as the web has been commercialised and *certainly* as long as it has been popular.
One could conceivably - just - make the argument that people who posted content to the web in its infancy did not realise that their content was going to be publically available and copied by others for the purposes of searching and/or archiving. It certainly wouldn't apply to anything published in (at least) the last ten years, however.
OK, show me ONE legal system that says copyright is terminated and no longer exist as soon as you publish something on the net! Just one is OK!!
Firstly, I never made any such claim. I merely said that publishing content on the public internet is done so with - at the very least - knowledge that it will be copied, indexed and made publically available without restriction and - more likely - published with the *intent* of that happening.
Secondly, it seems the ruling in this case has pretty much said what you're complaining about is legal (or, at worst, not illegal) - and a good thing, too.
Just because anyone can come and see your page or what you present on a web page (for as long as it is there only though), does not mean anyone in turn are allowed to start copying and distributing the content.
Well if they couldn't, both HTTP caches and indexing search engines (amongst numerous other fundamental aspects of the internet) would not only be illegal, but impossible. AFAIK, no court has ruled in this fashion.
There is no copyright law saying so either, feel free to check all of them.
You keep flicking back and forth between asking me what I think and asking me what the law says. Please make up your mind which you want to discuss.
Someone that show something on the net does NOT tell everyone to copy and distribute it in any way, well there are probably some pages that specifically states so but not in general.
Sorry, I cannot decipher what you mean.
Perhaps, then, you're not in the best position to comment on how responsive Windows developers are to feedback ?
When was the last time you sent a developer feedback about their Windows software ?
By this (ridiculous) logic, "very few people actually use" cars.
"Most popular" by what measure ?
So how was McDonalds responsible when the woman spilled the hot coffee in her lap, and M got sued for $1.0M. So now McDonalds have to put hot warnings on their coffee. Is is not reasonable to assume that unless someone is really "stupid", they would know that coffee is served hot.
This is OT, however, the reason McDonalds was found liable was because they were a) serving coffee at a far higher temperature than anyone would reasonably call "hot enough" and b) because they had received numerous complaints about the excessively high temperature their coffee was being served at - and other injuries it had caused - and done nothing about it.
The woman who has burned certainly shares some of the blame for clumsily opening the coffee in a rather inadvisable fashion, but what would have been a relatively minor burn and reminder that it was a silly thing to do, instead became a very serious and debilitating injury because the McDonald's coffee in question was so much hotter than expected.
There was clearly a valid case against McDonalds, in that instance. The problem was not that the coffee was hot, the problem that it was *unnecessarily* and *unexpectedly* hot, that McDonalds knew this and that they did nothing the remedy the situation.
I agree with you to a point. Sure, the user needs to be more informed on basic security practices and nothing can defeat a secure system better than a 'stupid user'. But, every product made to interact with human beings take this as a given: "Shit happens". In cars [...]
Here's the problem: computers aren't cars. Now, while the good old "car analogy" has some valid uses, this is not one of them. A car is a device manufactured, designed and optimised to do a fairly small set of well-defined tasks. Take the car outside of its "normal" parameters (eg: overloading, driving on ice with regular tyres, using the wrong type of fuel) and the car - and possibly the user - will be damaged. This is not a failure in the engineering of the car, it is a user error.
Computers are designed to handle, essentially, completely arbitrary, end-user defined tasks. Even when you throw an operating system into the mix, which reduces the set of possible tasks slightly, you *still* have a list of potential uses which is, for all intents and purposes, infinite.
Trying to design a "computer" (and by this I'm talking about the hardware+OS combination) so that it can only perform "good" tasks and not allow anything "bad" to happen is, essentially, impossible. As long as the design requirement of "arbitrary code execution" remains, the onus is always on the end user to decide which code is "good" and which code is "bad".
The same is and should be done in computers. User accounts are a must. Patching vulnerabilities should also not be mandatory. Even for the pirates.
Problem is most malware a) doesn't really require anything more than a standard user account and b) isn't exploiting software vulnerabilities (let alone unpatched ones). Which is not to say doing this things is a worthless endeavour, merely that it's nothing close to a "fix".
Microsoft has the power to enforce better security in windows by using better default configurations (your average user won't bother hardening his OS).
This is something they can - and are - doing with Vista. But they can't do it retroactively.
The other thing we need is diversity. Having one OS to rule them all is a vulnerability within itself, it just amplifies the damage.
Build it and they will come.
Security is a process that will never practically end, and blaming the user for "Almost all" malware infestations is to much ostrich-reasoning for my likings, not that it isn't true as things are today, especially on the windows side...
It's true on all platforms and it's not going to change as long as computers are able to run arbitrary code.
Yes. More accurately, the Doctor should not be charging so much in the first place.
did the husband break the law by stealing the medicie but leaving enough for base cost plus a little extra?
Yes. Note that this does not mean he did anything wrong.
You appear to be conflating "wrong" with "illegal" (your first question uses the term "wrong", your last the term "illegal"). "Illegal" does not imply "wrong" any more than "wrong" implies "illegal".
Somehow I doubt the silly woman who drove into the back of me a few weeks ago because she was busy talking on the phone and doing her nails will have much luck in court trying to blame Mazda for it.
The vast bulk of malware infestations are like the innatentive driver causing an accident, not the person being driven into whose car explodes.
Last I checked Microsoft *do* patch their software.
And if a particualr product is found to make it easier user-caused accidents (a la the tires on Ford Explorers), the company recalls them. Why is Windows lack of security exempt from this?
Because 99% of the time it's not lack of security in the OS, it's lack of common sense in the operator.
Of course, if you want to go down the path of "software liability", you can say goodbye to any meaningful amount of OSS acceptance.
Firstly, the "underlying problem" is almost always the user.
Secondly, anti-spyware and anti-virus software is protecting against things OS-level security *can't*.
Thirdly, as long as users are able to run arbitrary software, malware will always be a problem.