In France, there are rules preventing 18+ games from being sown on TV before 22:30. Even channels broadcasting on cable, satellite and dsl networks must respect those rules. That poses a problem to channels like Nolife TV, specialized in video games, because a lot of games get a PEGI 18 rating - if the player is able to kill a human-looking enemy, and this is done in a somewhat realist context, it's PEGI 18. As a result, they must cram discussion of a lot of games in a small time slot.
The rule was originally designed for movies, by the way, but the French movie rating is much more relaxed than the games rating. For example, the last James Bond movie did not get any restriction at all, it would be PEGI 18 if it were a game. But the movie rating boards in Europe use different standards.
At then end; it looks like Nintendo took the most restrictive of those rules, and applied them to everyone, as if the WiiU was a TV channel. This will hurt them in more liberal markets. It does not help that Nintendo of Europe is headquartered in Germany, which has the most extreme restrictions on video games, and still requires a separate, different, ugly, enormous, unremovable logo on game packaging and game disks. And this is after the PEGI rating board mainly standardized on rules very close to the German ones...
Just implement the calculation of any irrational number. If I remember well, you get an infinite string of digits, non-repeating, in any base you take.
The company claims it does it differently, with base stations using the ISM band to discuss directly with the devices on the field, and advanced signal processing in the base station to detect those signals.
Flash memory is nowadays only rated for 10 years retention. And that duration is valid only with a strong error correction code, and will be reduced by reading and writing. I really wonder how long today's devices will last.
The problem is the static leakage of transistors. It increases as the node width decreases, and for a given node you have two choices to generate a transistor: either high-speed and high-leakage, or low-speed and low-leakage. Even with DVFS enabled, you will have better power results if you use the CPU with the slow transistors than the one with the fast transistors. Hence the switching between two types of cores with different optimizations but executing the same code.
Conditional instructions are also available in the AArch64 assembler. But instead of the condition affecting whether the instruction is executed or not as in AArch32, it affects its behaviour. For example, you have the following Aarch64 instruction that can do what cmov does.
CSEL Wd, Wn, Wm, cond;Wd = if cond then Wn else Wm.
It's more efficient that the AArch32 case, where you have to use two instructions to achieve the same result (MOVcond followed by MOVcondbar)
Apple is rich enough to skip eMMC based memory for its iDevices, so it does not necessarily need this kind of file system. The NAND or eMMC trade-off is 'Spend (a lot of) money once to write your own FTL, and adapt it for each new chip' or 'Buy a chip with hardware FTL and a standard interface for a higher price".
You can check the tear-downs for all Apple devices: all of them directly use NAND, which makes sense. Apple buys large enough numbers of Flash to have reliable sources, and can invest the money to write a FTL.
The problem is that even with a translation layer for block access, flash-based devices have limitations, which means that different usage patterns can dramatically change the performance of the device.
For a (simplified) example, to write a file in ext3, you need to store the new data for the file, but you also need to store other metadata: the location of the data blocks themselves in the inode, the file size in the directory, the journaling data. This means that you have four 'internal block descriptors' open for writing at the same time.
But block descriptors are a limited resource in SSDs, and even more so for low-cost eMMC devices. This means that with only two or three open files with regular writing, you could quite easily lead to some kind of thrashing state, with the device quickly opening and closing descriptors. Since flash memory writing is strongly constrained, this means that a whole block (2 MiB block size is common) containing a descriptor will need to be erased before its next use. As a result, each block only contains little interesting data, and writing only a small amount of data leads to a lot of flash write and erase access. This problem is called write amplification, and reduces both the disk's performance and its durability.
The F2FS design is a log-based design, where all files on the disk share 6 common writing areas, for each kind of stored data, where the information is stored as it arrives. This will have a very positive effect against the write amplification problem, and is an example of how an adapted file system can have a positive impact, even on block-based devices.
Why do YOU lie ? On iOS, Apple is already the gatekeeper and you can't avoid it.
And from their current moves, it clearly looks like both Apple and Microsoft would do it on their desktop environments if they could get away with it. They cannot right now, and the warnings from Notch, Gabe Newell, etc. are clearly there to raise awareness on this issue, and try to foil the OS makers' strategy.
There was an other explanation for the very high price of recorded media in Japan. From what I understood, it was impossible to sell recorded media with rental restrictions, which meant that the price of a VHS/Laserdisc (at that time) was set to what rental shops could afford. As a result, the price was set much higher than in western countries, where you have different markets for rental and home video.
Valve does not prevent a developer from distributing games through any other mean, and there is no lack of concurrence in the Digital Game Delivery market. Self-publication is very cheap, and platforms like Steam are intended for developers that are ready to invest money to respect Valve's conditions, in exchange for an improved revenue through a better exposition to gamers that are used to buy their games.
This is quite different from the current Microsoft and Apple tactics of using their power as an OS provider to extract a "gatekeeper tax" on all programs sold for their platform.
It is legal to encrypt anything in France since 1996 for 128-bit symmetrical keys, and for any key since 2004. While the law was valid for a long time, I do not have knowledge of any prosecution on that basis.
There is a big effort in progress among well-known SoC manufacturers to work towards a unified ARM kernel, as a result of Linus Torvalds's crictics last year. Today, we see Texas Instruments, Freescale, ST/Ericsson and Samsung working to integrate their existing platforms into this new model, whereas the newer platforms - CSR's prima2 for example - are required to use all the new features: Device tree for ARM, the unified clock framework, pinmux, etc. must be used to get a new platform accepted in the mainline.
But for now, there is no sign of integrating the sunxi platform, supporting the Allwinner A10 chip, in the mainline. The available repository for the kernel is based on Linux 3.0.x with Android patches, before the implementation of many of these new features, and there were no attempts to submit the new architecture to the ARM Linux mailing list. This means that there is no chance of getting the A10 supported by Linus's tree before a long time.
Another problem is the sign-off for the commits that need to enter in the mainline. If the A10 support code was written by Allwinner employees, it will be necessary to get their sign-off, as it is required for any code integrated in the kernel nowadays. This means that no third party may commit their code until this is done, even if it is correctly marked as GPLv2. This was added in the wake of the SCO fiasco, to ensure that all source code lines could be attributed to their original writer, and thus have the author's word that it is not copied closed source code.
As long as all these issues are not resolved, the Allwinner based devices will remain second class citizens in the Linux world.
Samsung slavishly copied Apple and Nokia has clearly proven you can design a next-gen phone that looks nothing like the iPhone
Apple paid $1 billion to Nokia, plus $8 per device for patent infringement because iPhone's technical parts were infringing Nokia patents. And that is for real, hardware patents, not bouncing menus or pinch-to-zoom. Apple is as much guilty as Samsung may be, and I really wish Nokia went all the way and refused licensing instead, to give Apple a taste of its own medecine.
A single example: the "make xconfig" tool for the Linux kernel requires a specific variant of the QT toolkit to build, and the correct variant is QT3 for 2.6.32.x, while it is QT4 for the current dev branch. If the build environment points to the wrong QT version, the build will fail with cryptic messages, which means that you would need to sandbox your kernel compilation to compile both on the same machine. Of cource, most of the people in this situation uses the "make menuconfig" option, that uses libcurses in a terminal emulator instead.
I don't know whether this is a problem with the kernel's developpers or a general QT3/4 problem, but this is for sure a bad situation, and I expect that other projects have had the same kind of issues (forward & backward incompatibilities on the source level) with QT.
The "concatenated SMS" protocol, which removes this limit in exchange for a higher message count, existed before twitter started to be relevant: It is defined in GSM/3GPP release R99, which was already widely deployed even in low-cost devices in 2006, when twitter was created.
Do you realize that the chip on the other end of a SATA link - typically the controller in the SSD you're using right now - has a lot of chances to be an ARM chip ? It is the case for common SSD disk controllers (Marvell or Sandforce).
And even if it is not common in today's products, there are a lot of recent high-level ARM SoCs that offer SATA - not least because its low pin count makes it easier to route on the board in the end than a parallel bus. For example, TI's OMAP5, Freescale i.MX53 or CSR's Prima 2 have SATA support.
Given that UTC dates from 1961, and that the leap second concept comes from this, it rather means that a lot of coders are dealing with time without being careful, rather than a fault from the astronomers, who mostly use TAI (without the leap seconds) anyway.
Leap years, daylight saving time, local calendars, all those are also complex issues that need to be taken into account when writing software that needs to correctly handle time. And this is before taking into account the issues brought by the hardware, as clock drift, RTC limitations and integer wraparound, that need to be addressed by the OS developers.
All this means that when it comes to time, most developers should leave it to the OS and language libraries. And those implementing these libraries should take their work seriously, because time is a complex matter, as Y2K and the current issue remind us.
You realise that USB was an Intel standard that was pioneered exclusively on Macs, right?
To be precise, it was a Intel, Microsoft, Compaq, and NEC standard, as you can see in the USB 1 specification. You already had Compaq PCs under Windows 95 with USB installed.
Apple probably adopted it in 1998 because its proprietary ADB was completely outdated, Firewire was too expensive for cheap peripherals, and Macs did not have the market share to impose a new competing standard.
I've also wondered if the FRAND licenses held by component manufacturers like Qualcomm extend to Apple.
It does not. One of my previous employers tried to play this card with the MPEG-LA for digital TV decoders, and in the end they had to settle and pay for the MPEG2 patents. But Apple lawyers may be more skilled and success where others have failed.
If you're doing memory transfers to the world outside the CPU (DMA, alternate bus masters, peripherals, etc), you need to have the whole data flushed out of the cache, or the contents of the cache invalidated to avoid reading stale data. In many of those cases, the data must pass through the memory bus, and you cannot avoid cache misses. Hence the importance of low-latency memory access.
From the PDF, the implementation mistake is to give the attacker feedback on whether the tried key is correct after the first half of authentication (phase M4), and then after the complete authentication (phase M6). Since the PIN is only 8 digits, and the last one is a checksum, the problem is reduced to guessing 1 number in 10000, and then 1 in 1000.
The document states that there are few possible mitigations for the problem. However, it skips the obvious one: do not notify authentication success/failure until the response to the M6 message. This would restore the 1 in 10,000,000 guessing complexity of the PIN code, without changing the protocol. It should even be a new issue tested by the compliance suite the vendors need to pass to get the WPS certification.
Slashdot Mirror
In France, there are rules preventing 18+ games from being sown on TV before 22:30. Even channels broadcasting on cable, satellite and dsl networks must respect those rules. That poses a problem to channels like Nolife TV, specialized in video games, because a lot of games get a PEGI 18 rating - if the player is able to kill a human-looking enemy, and this is done in a somewhat realist context, it's PEGI 18. As a result, they must cram discussion of a lot of games in a small time slot.
The rule was originally designed for movies, by the way, but the French movie rating is much more relaxed than the games rating. For example, the last James Bond movie did not get any restriction at all, it would be PEGI 18 if it were a game. But the movie rating boards in Europe use different standards.
At then end; it looks like Nintendo took the most restrictive of those rules, and applied them to everyone, as if the WiiU was a TV channel. This will hurt them in more liberal markets. It does not help that Nintendo of Europe is headquartered in Germany, which has the most extreme restrictions on video games, and still requires a separate, different, ugly, enormous, unremovable logo on game packaging and game disks. And this is after the PEGI rating board mainly standardized on rules very close to the German ones...
Just implement the calculation of any irrational number. If I remember well, you get an infinite string of digits, non-repeating, in any base you take.
The company claims it does it differently, with base stations using the ISM band to discuss directly with the devices on the field, and advanced signal processing in the base station to detect those signals.
They have a pending patent on it, and they call it Forced Statistical FDMA
Flash memory is nowadays only rated for 10 years retention. And that duration is valid only with a strong error correction code, and will be reduced by reading and writing. I really wonder how long today's devices will last.
The problem is the static leakage of transistors. It increases as the node width decreases, and for a given node you have two choices to generate a transistor: either high-speed and high-leakage, or low-speed and low-leakage. Even with DVFS enabled, you will have better power results if you use the CPU with the slow transistors than the one with the fast transistors. Hence the switching between two types of cores with different optimizations but executing the same code.
Conditional instructions are also available in the AArch64 assembler. But instead of the condition affecting whether the instruction is executed or not as in AArch32, it affects its behaviour. For example, you have the following Aarch64 instruction that can do what cmov does.
;Wd = if cond then Wn else Wm.
CSEL Wd, Wn, Wm, cond
It's more efficient that the AArch32 case, where you have to use two instructions to achieve the same result (MOVcond followed by MOVcondbar)
Apple is rich enough to skip eMMC based memory for its iDevices, so it does not necessarily need this kind of file system. The NAND or eMMC trade-off is 'Spend (a lot of) money once to write your own FTL, and adapt it for each new chip' or 'Buy a chip with hardware FTL and a standard interface for a higher price".
You can check the tear-downs for all Apple devices: all of them directly use NAND, which makes sense. Apple buys large enough numbers of Flash to have reliable sources, and can invest the money to write a FTL.
The problem is that even with a translation layer for block access, flash-based devices have limitations, which means that different usage patterns can dramatically change the performance of the device.
For a (simplified) example, to write a file in ext3, you need to store the new data for the file, but you also need to store other metadata: the location of the data blocks themselves in the inode, the file size in the directory, the journaling data. This means that you have four 'internal block descriptors' open for writing at the same time.
But block descriptors are a limited resource in SSDs, and even more so for low-cost eMMC devices. This means that with only two or three open files with regular writing, you could quite easily lead to some kind of thrashing state, with the device quickly opening and closing descriptors. Since flash memory writing is strongly constrained, this means that a whole block (2 MiB block size is common) containing a descriptor will need to be erased before its next use. As a result, each block only contains little interesting data, and writing only a small amount of data leads to a lot of flash write and erase access. This problem is called write amplification, and reduces both the disk's performance and its durability.
The F2FS design is a log-based design, where all files on the disk share 6 common writing areas, for each kind of stored data, where the information is stored as it arrives. This will have a very positive effect against the write amplification problem, and is an example of how an adapted file system can have a positive impact, even on block-based devices.
Why do YOU lie ? On iOS, Apple is already the gatekeeper and you can't avoid it.
And from their current moves, it clearly looks like both Apple and Microsoft would do it on their desktop environments if they could get away with it. They cannot right now, and the warnings from Notch, Gabe Newell, etc. are clearly there to raise awareness on this issue, and try to foil the OS makers' strategy.
There was an other explanation for the very high price of recorded media in Japan. From what I understood, it was impossible to sell recorded media with rental restrictions, which meant that the price of a VHS/Laserdisc (at that time) was set to what rental shops could afford. As a result, the price was set much higher than in western countries, where you have different markets for rental and home video.
Valve does not prevent a developer from distributing games through any other mean, and there is no lack of concurrence in the Digital Game Delivery market. Self-publication is very cheap, and platforms like Steam are intended for developers that are ready to invest money to respect Valve's conditions, in exchange for an improved revenue through a better exposition to gamers that are used to buy their games.
This is quite different from the current Microsoft and Apple tactics of using their power as an OS provider to extract a "gatekeeper tax" on all programs sold for their platform.
It is legal to encrypt anything in France since 1996 for 128-bit symmetrical keys, and for any key since 2004. While the law was valid for a long time, I do not have knowledge of any prosecution on that basis.
There is a big effort in progress among well-known SoC manufacturers to work towards a unified ARM kernel, as a result of Linus Torvalds's crictics last year. Today, we see Texas Instruments, Freescale, ST/Ericsson and Samsung working to integrate their existing platforms into this new model, whereas the newer platforms - CSR's prima2 for example - are required to use all the new features: Device tree for ARM, the unified clock framework, pinmux, etc. must be used to get a new platform accepted in the mainline.
But for now, there is no sign of integrating the sunxi platform, supporting the Allwinner A10 chip, in the mainline. The available repository for the kernel is based on Linux 3.0.x with Android patches, before the implementation of many of these new features, and there were no attempts to submit the new architecture to the ARM Linux mailing list. This means that there is no chance of getting the A10 supported by Linus's tree before a long time.
Another problem is the sign-off for the commits that need to enter in the mainline. If the A10 support code was written by Allwinner employees, it will be necessary to get their sign-off, as it is required for any code integrated in the kernel nowadays. This means that no third party may commit their code until this is done, even if it is correctly marked as GPLv2. This was added in the wake of the SCO fiasco, to ensure that all source code lines could be attributed to their original writer, and thus have the author's word that it is not copied closed source code.
As long as all these issues are not resolved, the Allwinner based devices will remain second class citizens in the Linux world.
Samsung slavishly copied Apple and Nokia has clearly proven you can design a next-gen phone that looks nothing like the iPhone
Apple paid $1 billion to Nokia, plus $8 per device for patent infringement because iPhone's technical parts were infringing Nokia patents. And that is for real, hardware patents, not bouncing menus or pinch-to-zoom. Apple is as much guilty as Samsung may be, and I really wish Nokia went all the way and refused licensing instead, to give Apple a taste of its own medecine.
There may have been other reasons for this issue, as a misconfigured enviroment, but I can assure you the I have encountered this kind of problem...
A single example: the "make xconfig" tool for the Linux kernel requires a specific variant of the QT toolkit to build, and the correct variant is QT3 for 2.6.32.x, while it is QT4 for the current dev branch. If the build environment points to the wrong QT version, the build will fail with cryptic messages, which means that you would need to sandbox your kernel compilation to compile both on the same machine. Of cource, most of the people in this situation uses the "make menuconfig" option, that uses libcurses in a terminal emulator instead.
I don't know whether this is a problem with the kernel's developpers or a general QT3/4 problem, but this is for sure a bad situation, and I expect that other projects have had the same kind of issues (forward & backward incompatibilities on the source level) with QT.
The "concatenated SMS" protocol, which removes this limit in exchange for a higher message count, existed before twitter started to be relevant: It is defined in GSM/3GPP release R99, which was already widely deployed even in low-cost devices in 2006, when twitter was created.
Now all we need is Android running on SH3 and we'll have gone full circle.
The newest Renesas application chips are based on ARM, with a SH4 core only as a coprocessor, so this particular monster will probably never exist.
Do you realize that the chip on the other end of a SATA link - typically the controller in the SSD you're using right now - has a lot of chances to be an ARM chip ? It is the case for common SSD disk controllers (Marvell or Sandforce).
And even if it is not common in today's products, there are a lot of recent high-level ARM SoCs that offer SATA - not least because its low pin count makes it easier to route on the board in the end than a parallel bus. For example, TI's OMAP5, Freescale i.MX53 or CSR's Prima 2 have SATA support.
Given that UTC dates from 1961, and that the leap second concept comes from this, it rather means that a lot of coders are dealing with time without being careful, rather than a fault from the astronomers, who mostly use TAI (without the leap seconds) anyway.
Leap years, daylight saving time, local calendars, all those are also complex issues that need to be taken into account when writing software that needs to correctly handle time. And this is before taking into account the issues brought by the hardware, as clock drift, RTC limitations and integer wraparound, that need to be addressed by the OS developers.
All this means that when it comes to time, most developers should leave it to the OS and language libraries. And those implementing these libraries should take their work seriously, because time is a complex matter, as Y2K and the current issue remind us.
You realise that USB was an Intel standard that was pioneered exclusively on Macs, right?
To be precise, it was a Intel, Microsoft, Compaq, and NEC standard, as you can see in the USB 1 specification. You already had Compaq PCs under Windows 95 with USB installed.
Apple probably adopted it in 1998 because its proprietary ADB was completely outdated, Firewire was too expensive for cheap peripherals, and Macs did not have the market share to impose a new competing standard.
I've also wondered if the FRAND licenses held by component manufacturers like Qualcomm extend to Apple.
It does not. One of my previous employers tried to play this card with the MPEG-LA for digital TV decoders, and in the end they had to settle and pay for the MPEG2 patents. But Apple lawyers may be more skilled and success where others have failed.
If you're doing memory transfers to the world outside the CPU (DMA, alternate bus masters, peripherals, etc), you need to have the whole data flushed out of the cache, or the contents of the cache invalidated to avoid reading stale data. In many of those cases, the data must pass through the memory bus, and you cannot avoid cache misses. Hence the importance of low-latency memory access.
From the PDF, the implementation mistake is to give the attacker feedback on whether the tried key is correct after the first half of authentication (phase M4), and then after the complete authentication (phase M6). Since the PIN is only 8 digits, and the last one is a checksum, the problem is reduced to guessing 1 number in 10000, and then 1 in 1000.
The document states that there are few possible mitigations for the problem. However, it skips the obvious one: do not notify authentication success/failure until the response to the M6 message. This would restore the 1 in 10,000,000 guessing complexity of the PIN code, without changing the protocol. It should even be a new issue tested by the compliance suite the vendors need to pass to get the WPS certification.
This is the idea behiind Ribbon Hero, a game designed to learn to use the ribbon interface of Microsoft Office.