Open spectrum advocates can gain a lot of credibility by demonstrating techniques like SDR, cognitive radio, and mesh networking in the existing unlicensed bands. (The article mentions LocustWorld, which is a commendable example.) Once there's quantitative information on the benefit of the technology it will be appropriate to ask the FCC to reconsider the current spectrum policy.
Yeah, I think they should have explained that. Presumably a Web graph is what you get when you treat each URL as a node and each link as an edge in a graph. PageRank is an algorithm used by Google that takes a Web graph as input.
The old Prism II chipset does not and cannot support 802.11g. There's a new Prism GT that supports 802.11g, but of course it has no docs or Linux drivers.
You're confusing channel access schemes (FDMA, TDMA, CDMA, CSMA, etc.) with modulations (PSK, CCK, PBCC, QAM, OFDM, etc.). And no doubt I'm also confusing something, since I'm not a wireless engineer.
OFDM is an encoding, not a protocol. Both Wi-Fi and WiMax (802.16) use OFDM, and I wouldn't be surprised if 4G (802.20) systems end up using it as well.
The sad part is that with IBM's lack of progression of the chip, the PPC has not maintained the competitive advances it SHOULD have been implemented years ago that it is capable of doing.
PPC processors have a 68K compatibility mode. Basically, the instructions can have a 68K flag that tells the processor to execute as such. The problem is that 68K code and PPC code aren't even remotely related to one another, and the processor executes 68K code much more slowly. The processor incurs a performance hit every time it switches mode in either direction, and if you are multitasking between a PPC and a 68K app, you can incur that penalty several every single instruction if you're unlucky.
DDR and QDR techniques are not overclocking; a 4x200MHz bus produces virtually the same bandwidth as an 800MHz bus would. (Notice how these systems have over 5GB/s of memory bandwidth.)
AMD is already ahead of Intel with on-chip dual channel memory controllers and HyperTransport.
A company like VMWare could implement a psuedo-driver that appeared to the operating system to be real Palladium hardware. Of course this driver wouldn't have access to the keys inside your real Palladium hardware, but it could generate its own keys. Other machines on the internet would never be able to tell the difference.
Not quite. The real Palladium hardware has a certificate issued by the manufacturer (whose certificate is issued by MS, etc.), which other machines can verify.
I happened to have dinner with Alan Cox the other night and I couldn't resist asking about this; he said he only gets mistaken for the other Alan Cox about once every 6 months.
The answer to that is in the article (sort of). People don't like things that look alike but work differently. You can make KDE and GNOME apps look alike, but a theme won't make them work alike.
That happens with pretty much all MPEG-1 files and it's a real PITA. Since the bug has been there for years I don't have any hope of it ever being fixed.
Slashdot Mirror
Open spectrum advocates can gain a lot of credibility by demonstrating techniques like SDR, cognitive radio, and mesh networking in the existing unlicensed bands. (The article mentions LocustWorld, which is a commendable example.) Once there's quantitative information on the benefit of the technology it will be appropriate to ask the FCC to reconsider the current spectrum policy.
BitTorrent can distribute directories of files, so why add the extra obfuscation?
We use EVMS on a 2.6TB file server and it works fine.
There's only so much storage you can attach to a single RAID controller; LVM allows you to combine multiple RAID controllers.
LVM also allows you to dynamically add and remove storage to a single filesystem.
Also, they apparently don't understand the difference between (C) and TM, or the first thing about trademark law.
Yeah, I think they should have explained that. Presumably a Web graph is what you get when you treat each URL as a node and each link as an edge in a graph. PageRank is an algorithm used by Google that takes a Web graph as input.
The old Prism II chipset does not and cannot support 802.11g. There's a new Prism GT that supports 802.11g, but of course it has no docs or Linux drivers.
The 5GHz U-NII band is 3X the size of the 2.4GHz ISM band (IIRC), but it's hardly used at all. Is there demand for even more free spectrum?
You're confusing channel access schemes (FDMA, TDMA, CDMA, CSMA, etc.) with modulations (PSK, CCK, PBCC, QAM, OFDM, etc.). And no doubt I'm also confusing something, since I'm not a wireless engineer.
OFDM is an encoding, not a protocol. Both Wi-Fi and WiMax (802.16) use OFDM, and I wouldn't be surprised if 4G (802.20) systems end up using it as well.
...Theora or H.264?
The sad part is that with IBM's lack of progression of the chip, the PPC has not maintained the competitive advances it SHOULD have been implemented years ago that it is capable of doing.
Lack of progression? You mean like #1 in SPEC FP 2000, #4 in SPEC INT 2000 and #2 in TPC-C? IBM has been doing a great job with PowerPCs, but they've been concentrating on server versions.
PPC processors have a 68K compatibility mode. Basically, the instructions can have a 68K flag that tells the processor to execute as such. The problem is that 68K code and PPC code aren't even remotely related to one another, and the processor executes 68K code much more slowly. The processor incurs a performance hit every time it switches mode in either direction, and if you are multitasking between a PPC and a 68K app, you can incur that penalty several every single instruction if you're unlucky.
Wrong, wrong, wrong. Where did you get that from?
DDR and QDR techniques are not overclocking; a 4x200MHz bus produces virtually the same bandwidth as an 800MHz bus would. (Notice how these systems have over 5GB/s of memory bandwidth.)
AMD is already ahead of Intel with on-chip dual channel memory controllers and HyperTransport.
IIRC, all the computers on earth couldn't brute-force a large RSA key in our lifetime.
A company like VMWare could implement a psuedo-driver that appeared to the operating system to be real Palladium hardware. Of course this driver wouldn't have access to the keys inside your real Palladium hardware, but it could generate its own keys. Other machines on the internet would never be able to tell the difference.
Not quite. The real Palladium hardware has a certificate issued by the manufacturer (whose certificate is issued by MS, etc.), which other machines can verify.
I happened to have dinner with Alan Cox the other night and I couldn't resist asking about this; he said he only gets mistaken for the other Alan Cox about once every 6 months.
That's SACK and it's a good idea but a lot of systems still don't implement it.
You joke, but SPEC is a consortium made up of IBM, Intel, Sun, HP, etc.
The 970 does not support HyperTransport.
I believe it really is 105 degrees C; a lot of embedded components are rated for "industrial" temperature ranges of 100C.
They're not technically overclocked; the 7455 is rated 1GHz at 100C and 1.4GHz at 60C. Presumably the 1.3GHz@100C 7457 can be run near 2GHz at 60C.
The answer to that is in the article (sort of). People don't like things that look alike but work differently. You can make KDE and GNOME apps look alike, but a theme won't make them work alike.
That happens with pretty much all MPEG-1 files and it's a real PITA. Since the bug has been there for years I don't have any hope of it ever being fixed.
Great. They add support for upload rate limiting, and don't even bother to put support into the GUI for the OS X version.
It's in there; just keep looking.