Why would you spend $1200 or even $200 for an International Driving Permit? According to the Danish consolate International Driving Permits are available at most Danish police stations for a very modest 25 Krone (about $3 US).
Since you mention that you are from Denmark you presumably know this - I suspect there must be some other unmentioned fact that accounts for your willingness to pay such exorbitant rates for an IDP.
Are you perhaps trying to get an International Drivers Permit issued by a country other than Denmark because as a citizen of Denmark you are currently unable to get a license to drive in Denmark? A foreign IDP wouldn't be valid unless you also had a foreign license since IDP's are only valid when accompanied by a driver's license issued by the same country as the issuer of the IDP. Even if you did obtain a foreign IDP and a foreign driver's license, I don't think that would help. As I understand it, permanent residents of Denmark are required to obtain a Danish driver's license, they can't use an IDP.
The best defense is a good offense. It appears that the general practice of the U.S. military is to first achieve air superiority before moving in ground forces (who are presumably the primary users of the GPS system). Since a jammer is basically a big radio transmitter, turning a jammer on in the presence of one of these (also here) when the U.S. has air superiority is like announcing "Here are my precise coordinates, please shoot me with a missile."
I'll confess that I agree that my predictions don't contradict any of the significant conclusions of the article, but I can't help it if Leon Erlanger is right.:)
The article does a good job of laying out the merits of each of these interconnection technologies but it does carefully avoid actually picking winners and losers.
I'll stick my foot out and make some predictions regarding the various interconnection technologies for desktop and server machines.
PCI and PCI-X will be replaced by 3GIO as the predominant expansion bus.
Proprietary cpu-to-chip busses like Intel's Pentium 4 bus, AMD's EV6 bus or Motorola's MaxBus will, in the longer term, be eventually replaced by 3GIO for I/O (with the memory controller a part of the CPU). In the shorter term AMD and maybe the PowerPC camp will make CPUs with HyperTransport. An interesting possibility for PowerPC CPUs might be Motorola choosing to make PowerPC CPUs with a RapidIO bus for the embedded market and Intel choosing 3GIO (or perhaps HyperTransport).
Chip-to-chip interconnection will standardize around either HyperTransport or 3GIO depending on the flavor of CPU and expansion bus in the machine. It seems awkward to use HyperTransport to connect a 3GIO expansion bus, but it makes a lot of sense to use HyperTransport to connect a PCI expansion bus.
The MIPS and PowerPC camps will push RapidIO in the embedded market, but I foresee RapidIO eventually being displaced by 3GIO in mid-range applications like printers.
In the server-to-server and server-to-storage applications where IP networking is not appropriate FibreChannel will eventually be replaced by InfiniBand.
Does anyone know a major site that actually _uses_ spews? I couldn't find one.
I couldn't find any statements (definitive or otherwise) from any big players saying that they are using SPEWS. However, by looking at the reports in news.admin.net-abuse.email and the bounce messages that I asked to be forwarded to me I think the two largest users of SPEWS are:
Pacific Bell - a large telecom on the US west coast.
Outblaze - a mailbox outsource company which handles mail for such sites as Mail.Com (a free mailbox provider).
FireWire chips just like all other periphrials are hooked up to the Southbridge of the fucking memory controller. The limits of the SB are the limits of the periphrial interface for the computer. Since the SB is running at 33MHz with either a 32 or 64 bit bus your onboard FW chip isn't going to have any better throughput than a periphrial card.
If we are talking about the Macintosh, I disagree. As I mentioned earlier in this discussion, the current line of G4 desktop machines use an Apple custom north and south bridge chipset. The north bridge (called Uni-N) is the memory controller and bridge to the AGP and PCI busses. Unlike the design of PC chip sets, the north bridge also contains an integrated controller for both FireWire and Gigabit Ethernet. Apple has does this specifically to get around the performance limitations of the PCI bus (and to reduce chip count). Less demanding IO hangs off the south bridge.
Gigabit Ethernet is close to the theoretical maximum bandwidth for 32-bit 33-MHz PCI. IEEE-1394b is more than triple that speed, which would put it well beyond what ordinary PCI can deliver. Does this mean that this new flavor of FireWire will require 64-bit 66-MHz PCI (rarely seen outside servers and workstations), or will it sit on some other bus (such as HyperTransport)?
Probably neither. The current line of G4 desktop machines use an Apple custom north and south bridge chipset. The north bridge (called Uni-N) is the memory controller and bridge to the AGP and PCI busses. More significantly, it also contains an integrated controller for both FireWire and Gigabit Ethernet. The lower speed I/O interfaces (such as USB) are handled in the standard way, i.e. by the south bridge. To keep down both the chip count down as well as costs I expect that Apple will continue with their current strategy of integrating the high performance IO controllers directly into the design of their north bridge.
Apple, however, has expressed interest in HyperTransport. Apple might choose to integrate the PCI controller into the south bridge and connect the north bridge and south bridge by HyperTransport instead of by PCI. Furthermore, if Motorola or IBM should decide to release a PowerPC chip with a HyperTransport interface you will likely see Apple switch from the Max bus to the HyperTransport bus to connect the north bridge to the CPU.
I don't think the weight of the plane was as signficant as one might think. The number of people in the WTC reaches about 25,000 at the daily peak, and at the time of the attack the number of people has been estimated as 10,000. Just the weight of the additional 15,000 people who would normally have arrived later in the day could conservatively be estimated as at least 1,500,000 pounds, so I don't find the weight of a 400,000 pound aircraft as dramatic as it sounds.
One idea I've seen is to encode a fingerprint in the form of alternate (yet correct) spellings sprinkled throughout the text.
A straightforward application of this can be defeated through the collusion of two purchasers of the book. The words with the altered spelling can be identified by comparing the slightly different copies of the book. These words can then be either modified to use a standard spelling. Or for that matter, they could be modified to use a mixed set from the the two books - thereby making the new copy look like it was purchased by someone other than either of the two coluding purchasers.
In fact, what you describe is essentially a form of watermarking, and the recent defeat of SDMI by a team of researchers from Princeton University shows that even sophisticated watermarking techniques can be broken.
Actually this is strange. They did a pumpkin launching contest (rugby ball throwing in the UK), and in the UK, one team used an air pressure gun. The other team used a catapult / slingshot device. When I saw the US version of the challenge, again, one team used a very similar looking (and functioning) air pressure cannon, and the other team used a catapult. I didnt know what conclusion to come to about this.
The thing you have to remember is that the purpose of the show is not the contest, the purpose is to teach science.
The approach the competing teams use to attack the challenge is usually heavily influenced by the skills of the expert provided by the producers as the 4th member of each team. In order to enhance the instructive nature of the show it is not uncommon for the producers to deliberately choose experts with different skill sets that the producers feel will encourage the teams to choose methods that illustrate different scientific principles.
Slashdot Mirror
It isn't that no one reads the Los Angeles Times, it's that no one reads the Los Angeles Times Magazine .
Why would you spend $1200 or even $200 for an International Driving Permit? According to the Danish consolate International Driving Permits are available at most Danish police stations for a very modest 25 Krone (about $3 US).
Since you mention that you are from Denmark you presumably know this - I suspect there must be some other unmentioned fact that accounts for your willingness to pay such exorbitant rates for an IDP.
Are you perhaps trying to get an International Drivers Permit issued by a country other than Denmark because as a citizen of Denmark you are currently unable to get a license to drive in Denmark? A foreign IDP wouldn't be valid unless you also had a foreign license since IDP's are only valid when accompanied by a driver's license issued by the same country as the issuer of the IDP. Even if you did obtain a foreign IDP and a foreign driver's license, I don't think that would help. As I understand it, permanent residents of Denmark are required to obtain a Danish driver's license, they can't use an IDP.
The best defense is a good offense. It appears that the general practice of the U.S. military is to first achieve air superiority before moving in ground forces (who are presumably the primary users of the GPS system). Since a jammer is basically a big radio transmitter, turning a jammer on in the presence of one of these (also here) when the U.S. has air superiority is like announcing "Here are my precise coordinates, please shoot me with a missile."
I'll confess that I agree that my predictions don't contradict any of the significant conclusions of the article, but I can't help it if Leon Erlanger is right. :)
The article does a good job of laying out the merits of each of these interconnection technologies but it does carefully avoid actually picking winners and losers.
I'll stick my foot out and make some predictions regarding the various interconnection technologies for desktop and server machines.
PCI and PCI-X will be replaced by 3GIO as the predominant expansion bus.
Proprietary cpu-to-chip busses like Intel's Pentium 4 bus, AMD's EV6 bus or Motorola's MaxBus will, in the longer term, be eventually replaced by 3GIO for I/O (with the memory controller a part of the CPU). In the shorter term AMD and maybe the PowerPC camp will make CPUs with HyperTransport. An interesting possibility for PowerPC CPUs might be Motorola choosing to make PowerPC CPUs with a RapidIO bus for the embedded market and Intel choosing 3GIO (or perhaps HyperTransport).
Chip-to-chip interconnection will standardize around either HyperTransport or 3GIO depending on the flavor of CPU and expansion bus in the machine. It seems awkward to use HyperTransport to connect a 3GIO expansion bus, but it makes a lot of sense to use HyperTransport to connect a PCI expansion bus.
The MIPS and PowerPC camps will push RapidIO in the embedded market, but I foresee RapidIO eventually being displaced by 3GIO in mid-range applications like printers.
In the server-to-server and server-to-storage applications where IP networking is not appropriate FibreChannel will eventually be replaced by InfiniBand.
Perhaps taking it with a pillar of salt would be more appropriate?
I couldn't find any statements (definitive or otherwise) from any big players saying that they are using SPEWS. However, by looking at the reports in news.admin.net-abuse.email and the bounce messages that I asked to be forwarded to me I think the two largest users of SPEWS are:
Pacific Bell - a large telecom on the US west coast.
Outblaze - a mailbox outsource company which handles mail for such sites as Mail.Com (a free mailbox provider).
If we are talking about the Macintosh, I disagree. As I mentioned earlier in this discussion, the current line of G4 desktop machines use an Apple custom north and south bridge chipset. The north bridge (called Uni-N) is the memory controller and bridge to the AGP and PCI busses. Unlike the design of PC chip sets, the north bridge also contains an integrated controller for both FireWire and Gigabit Ethernet. Apple has does this specifically to get around the performance limitations of the PCI bus (and to reduce chip count). Less demanding IO hangs off the south bridge.
Probably neither. The current line of G4 desktop machines use an Apple custom north and south bridge chipset. The north bridge (called Uni-N) is the memory controller and bridge to the AGP and PCI busses. More significantly, it also contains an integrated controller for both FireWire and Gigabit Ethernet. The lower speed I/O interfaces (such as USB) are handled in the standard way, i.e. by the south bridge. To keep down both the chip count down as well as costs I expect that Apple will continue with their current strategy of integrating the high performance IO controllers directly into the design of their north bridge.
Apple, however, has expressed interest in HyperTransport. Apple might choose to integrate the PCI controller into the south bridge and connect the north bridge and south bridge by HyperTransport instead of by PCI. Furthermore, if Motorola or IBM should decide to release a PowerPC chip with a HyperTransport interface you will likely see Apple switch from the Max bus to the HyperTransport bus to connect the north bridge to the CPU.
I don't think the weight of the plane was as signficant as one might think. The number of people in the WTC reaches about 25,000 at the daily peak, and at the time of the attack the number of people has been estimated as 10,000. Just the weight of the additional 15,000 people who would normally have arrived later in the day could conservatively be estimated as at least 1,500,000 pounds, so I don't find the weight of a 400,000 pound aircraft as dramatic as it sounds.
For what it is worth, I recall an in the August 2000 Dr. Dobb's Journal in which a Microsoft engineer mentions that NVIDIA is planning on supporting OpenGL on the Xbox.
I haven't heard anything directly from NVIDIA about this, however.
In fact, what you describe is essentially a form of watermarking, and the recent defeat of SDMI by a team of researchers from Princeton University shows that even sophisticated watermarking techniques can be broken.
The thing you have to remember is that the purpose of the show is not the contest, the purpose is to teach science.
The approach the competing teams use to attack the challenge is usually heavily influenced by the skills of the expert provided by the producers as the 4th member of each team. In order to enhance the instructive nature of the show it is not uncommon for the producers to deliberately choose experts with different skill sets that the producers feel will encourage the teams to choose methods that illustrate different scientific principles.