it's patenting accurate thumbnails of virtual desktops and using those to swich between the desktop (as previews)
This sounds like an obvious extension to the virtual desktop pager used by several window managers already ten years ago. Sure, they didn't actually keep a scaled down picture of the window with contents, but rather a proportionally accurate gray rectangle for each window. Still, you are not supposed to be able to patent obvious stuff, and using a scaled image of the window instead of a gray box is certainly obvious.
If you free an object that completely covers a page, it is quite possible the malloc code will return it to the system
If it does, how? AFAIK, there are two ways to allocate memory and return memory, and those are the brk() and sbrk() functions, as well as the mmap() and the munmap() functions. Could those be mixed? I.e. could you sbrk(-x) an area that was mmap():ed? I think I read somewhere that process termination produces an implicit munmap on the entire process address space, implying that munmap() could be used on memory areas produced by brk().
Then what system call is used to allocate memory somewhere else than at the end of the data segment? Okay, I realize that large allocations are actually handled by mmap(2) (while smaller allocations are usually handled by brk(2) and sbrk(3)), but where is that memory placed in the process address space?
C++ has everything to do with it. Since C++ allows programmers to use actual pointers, it cannot run an automatic garbage collector and memory compactor in the background. This is because it would have to update pointers referring to data areas moved during compaction to reflect the new location, but it wouldn't know what memory elements are actually pointers, which makes this impossible.
This is why Java-style (C#-style, Python-style, whatever) references are good. They don't point to any real memory, but rather into a list of objects, where the real pointers are located. When compacting memory, only this list would have to be updated, as no data contains real pointers that would have to be updated and all pointers are available in this list.
The memory unused is too small to return to the OS
Actually, you can return any amount of memory to the OS, there is no smallest size. The problem is that the returned memory must be located at the end of the allocated memory (i.e. the end of the data segment), as memory is only returned to the OS by decreasing the data segment size (at least on *nix, the actual system call used by the malloc() family of functions to allocate more memory is brk(), which sets a new data segment size). If used data is stored after a section of unused memory, the unused memory cannot be returned regardless of size. If you would want to return that memory to the OS, you would have to move the used data to some lower part of the data segment, so that the unused memory is placed last. Since the used data probably contains pointers, this is not feasible to do, since they would have to be identified and updated to reflect the new location.
Any time Linux or OSX came out with a release, we heard all about how Longhorn would do the same thing only better. Of course, when the time came, none of those features were delivered.
Sounds like the old Microsoft Cairo project. Each time a competitor was about to release a new product or new version of a product, Microsoft would launch a press release stating how much better everything would be with Cairo, who would be just six months away. The press and potential customers turned away from the competitor and started to talk about the marvelous Cairo future instead.
Once you toss PHP into the mix, Linux is a very insecure operating system. I'll take FreeBSD any day.
Can you explain why Linux becomes a very insecure operating system with the addition of PHP, while FreeBSD with PHP is still a secure operating system (which is implied in your post)?
There is nothing wrong with using a SUV or a pickup truck for those purposes where such cars are needed. The problem is with all those that use them for nothing more than commuting or going shopping, when a normal car would do just as fine.
I don't even own a car. When I need one, I borrow my parents' car, a Volvo S60, but that doesn't happen too often. For most of my transportation purposes, I use the underground, which passes less than 100 yards from my living room window.
it's highly unlikely that someone living in south Texas will need much snow practice.
People never drive out of the state? People in Texas (or any other warm state) never go into the Rocky Mountains to go skiing? In those cases, some experience of driving in slippery conditions might be a good thing.
Sadly, no. That would be a really great idea, but here in the States, driving is seen as a right, not a privilege.
That's a pity. I think that slippery driving courses at least make people aware of the problem, hopefully driving more carefully in these conditions. It is surely no be-all-end-all solution, but it is certainly one step in the right direction.
One argument against the "driving is a right, not a privilege" thinking is that sure, you can drive all you want with no driving license whatsoever on your own land. Using the public roads OTOH, is a privilege that requires a certified skill level to acquire. This is because on the public road, you are not alone. By driving without the necessary skills on a public road, you do not only put yourself but other people at risk. And I would include slippery conditions as a requirement in any state that has even a slight risk of getting these conditions.
Here, they don't even bother shutting the schools down unless it looks like there's going to be at least half a foot.
I cannot even remember the schools being shutdown because of snow at all, and we sometimes get upwards of a foot of snow in the winter, at least on some winters. Most winters is more like at most 10 cm of snow.
In PA, a light dusting of snow results in car accidents and other hilarity.
Sadly, we get that here too, at least with the first snow each winter. Then a lot of cars are involved in accidents, either with other cars or just slipping off the road. It's like people forget that snow falls (almost) every winter. But then slippery driving courses have only been mandatory to get a license for some 15 years. Before that, people were only required to be able to drive in summer conditions to get a license.
We still see SUVs stuck in the ditch from time to time. Never cars, always SUVs.
This clearly sounds like the SUV drivers think they are invulnerable just because they have 4WD or maybe even because the car (a SUV is a car) is so big (it might feel unstoppable).
No, but where I live (Sweden) RWD cars are almost oddities nowadays. Few new cars use RWD here. Of course, those that still do mostly put the engine in the front, with well-known results such as the inability to climb any even slightly slippery incline.
Cars with RWD and the engine in the trunk should not have these problems, at least not to the same extent. The engine is heavy, and by putting it at the same end as the driving wheels, traction is substantially increased.
why don't we mandate that all agencies that receive money from Congress must not use cars with a MPG below 35? This includes charities, police departments, the Military, and even foreign governments.
I'm pretty sure that the M1 Abrams does not reach 35 MPG. Does that mean that it should be banned? Or do you have a different standard for tanks?:)
If you know how to drive with your car in all conditions
In the US, is it required to take a slippery driving course to get a drivers' license? Or is it up to the individual states? Anyway, it is mandatory here (in Sweden)since some time ago, and it was quite interesting. You learned what to do and what to not do, as well as what happens if you do the wrong thing (such as turning and braking at the same time).
How about studded tires? They are pretty common here in Sweden (I almost never see real snow chains), and they work well, although they do increase the wear and tear on the pavement.
I don't have any idea why so many Americans think they need 4WD SUVs in any climate where snowflakes are ever seen.
I think that it could be because historically, (IIRC) most American cars have been RWD. And anyone familiar with such know that they are not fun to drive in slippery road conditions. Thus they think that 4WD is the only answer, failing to realize that FWD is almost as good.
Except for people with rear-wheel-drive cars with the engine in the front, like the Volvo 240. You sometimes see them getting stuck at the bottom of a hill, since they cannot get enough traction to drive up the (icy) slope. But they really only have themselves to blame, since few cars use that configuration anymore.
Web servers "guaranteed" to work with IE make a lot of cash for Microsoft.
So if I would code a crappy (according to IE) web page, IIS would automatically change the code to work well with IE? No, I'm not asking you, I'm asking your PHB.
I don't think so. A government-mandated signature system would probably rather use PKI. Your key would need a(n expensive) signature from Verisign or another "trusted" signer. That would also scare away hobby programmers, and leave programming an exclusive domain of major companies, which we all know never make mistakes, and when they do, they can afford a few millions in fines.
Slashdot Mirror
Michael Dell?
Until Microsoft substantiates its allegations, those allegations can be dismissed as lies.
This sounds like an obvious extension to the virtual desktop pager used by several window managers already ten years ago. Sure, they didn't actually keep a scaled down picture of the window with contents, but rather a proportionally accurate gray rectangle for each window. Still, you are not supposed to be able to patent obvious stuff, and using a scaled image of the window instead of a gray box is certainly obvious.
You mean like a virtual desktop pager? I had one in Fvwm2 almost a decade ago.
If it does, how? AFAIK, there are two ways to allocate memory and return memory, and those are the brk() and sbrk() functions, as well as the mmap() and the munmap() functions. Could those be mixed? I.e. could you sbrk(-x) an area that was mmap():ed? I think I read somewhere that process termination produces an implicit munmap on the entire process address space, implying that munmap() could be used on memory areas produced by brk().
Then what system call is used to allocate memory somewhere else than at the end of the data segment? Okay, I realize that large allocations are actually handled by mmap(2) (while smaller allocations are usually handled by brk(2) and sbrk(3)), but where is that memory placed in the process address space?
C++ has everything to do with it. Since C++ allows programmers to use actual pointers, it cannot run an automatic garbage collector and memory compactor in the background. This is because it would have to update pointers referring to data areas moved during compaction to reflect the new location, but it wouldn't know what memory elements are actually pointers, which makes this impossible.
This is why Java-style (C#-style, Python-style, whatever) references are good. They don't point to any real memory, but rather into a list of objects, where the real pointers are located. When compacting memory, only this list would have to be updated, as no data contains real pointers that would have to be updated and all pointers are available in this list.
Actually, you can return any amount of memory to the OS, there is no smallest size. The problem is that the returned memory must be located at the end of the allocated memory (i.e. the end of the data segment), as memory is only returned to the OS by decreasing the data segment size (at least on *nix, the actual system call used by the malloc() family of functions to allocate more memory is brk(), which sets a new data segment size). If used data is stored after a section of unused memory, the unused memory cannot be returned regardless of size. If you would want to return that memory to the OS, you would have to move the used data to some lower part of the data segment, so that the unused memory is placed last. Since the used data probably contains pointers, this is not feasible to do, since they would have to be identified and updated to reflect the new location.
Sounds like the old Microsoft Cairo project. Each time a competitor was about to release a new product or new version of a product, Microsoft would launch a press release stating how much better everything would be with Cairo, who would be just six months away. The press and potential customers turned away from the competitor and started to talk about the marvelous Cairo future instead.
Except that Cairo never materialized.
Seems like it works again now. At least I get the predictably jumbled mess that Firefox 2 generates.
Multiply the time between vulnerability discovery and patch with a severity multiplier, and you may get a better metric.
You must be new here. :)
This is a very old tactic by Microsoft supporters to make Windows look much more secure than Linux.
Can you explain why Linux becomes a very insecure operating system with the addition of PHP, while FreeBSD with PHP is still a secure operating system (which is implied in your post)?
There is nothing wrong with using a SUV or a pickup truck for those purposes where such cars are needed. The problem is with all those that use them for nothing more than commuting or going shopping, when a normal car would do just as fine.
I don't even own a car. When I need one, I borrow my parents' car, a Volvo S60, but that doesn't happen too often. For most of my transportation purposes, I use the underground, which passes less than 100 yards from my living room window.
People never drive out of the state? People in Texas (or any other warm state) never go into the Rocky Mountains to go skiing? In those cases, some experience of driving in slippery conditions might be a good thing.
That's a pity. I think that slippery driving courses at least make people aware of the problem, hopefully driving more carefully in these conditions. It is surely no be-all-end-all solution, but it is certainly one step in the right direction.
One argument against the "driving is a right, not a privilege" thinking is that sure, you can drive all you want with no driving license whatsoever on your own land. Using the public roads OTOH, is a privilege that requires a certified skill level to acquire. This is because on the public road, you are not alone. By driving without the necessary skills on a public road, you do not only put yourself but other people at risk. And I would include slippery conditions as a requirement in any state that has even a slight risk of getting these conditions.
Here, they don't even bother shutting the schools down unless it looks like there's going to be at least half a foot.I cannot even remember the schools being shutdown because of snow at all, and we sometimes get upwards of a foot of snow in the winter, at least on some winters. Most winters is more like at most 10 cm of snow.
In PA, a light dusting of snow results in car accidents and other hilarity.Sadly, we get that here too, at least with the first snow each winter. Then a lot of cars are involved in accidents, either with other cars or just slipping off the road. It's like people forget that snow falls (almost) every winter. But then slippery driving courses have only been mandatory to get a license for some 15 years. Before that, people were only required to be able to drive in summer conditions to get a license.
This clearly sounds like the SUV drivers think they are invulnerable just because they have 4WD or maybe even because the car (a SUV is a car) is so big (it might feel unstoppable).
No, but where I live (Sweden) RWD cars are almost oddities nowadays. Few new cars use RWD here. Of course, those that still do mostly put the engine in the front, with well-known results such as the inability to climb any even slightly slippery incline.
Cars with RWD and the engine in the trunk should not have these problems, at least not to the same extent. The engine is heavy, and by putting it at the same end as the driving wheels, traction is substantially increased.
I'm pretty sure that the M1 Abrams does not reach 35 MPG. Does that mean that it should be banned? Or do you have a different standard for tanks? :)
In the US, is it required to take a slippery driving course to get a drivers' license? Or is it up to the individual states? Anyway, it is mandatory here (in Sweden)since some time ago, and it was quite interesting. You learned what to do and what to not do, as well as what happens if you do the wrong thing (such as turning and braking at the same time).
How about studded tires? They are pretty common here in Sweden (I almost never see real snow chains), and they work well, although they do increase the wear and tear on the pavement.
I think that it could be because historically, (IIRC) most American cars have been RWD. And anyone familiar with such know that they are not fun to drive in slippery road conditions. Thus they think that 4WD is the only answer, failing to realize that FWD is almost as good.
Except for people with rear-wheel-drive cars with the engine in the front, like the Volvo 240. You sometimes see them getting stuck at the bottom of a hill, since they cannot get enough traction to drive up the (icy) slope. But they really only have themselves to blame, since few cars use that configuration anymore.
So if I would code a crappy (according to IE) web page, IIS would automatically change the code to work well with IE? No, I'm not asking you, I'm asking your PHB.
I don't think so. A government-mandated signature system would probably rather use PKI. Your key would need a(n expensive) signature from Verisign or another "trusted" signer. That would also scare away hobby programmers, and leave programming an exclusive domain of major companies, which we all know never make mistakes, and when they do, they can afford a few millions in fines.