Modern cellphones adjust the transmit power based on the Signal-Noise Ration. If there are a lot of cellphones around, then the noise is high, so the phone has to increase the transmit power (reducing battery life). If the cell phone is able to find a frequency with very little noise, it can transmit with very little power, and still have the same quality. If a software radio is used, the number of possible frequencies becomes much larger. The gains from using this method should more than offset the losses from the slightly less efficient analog stages and higher-speed digital processors and converters.
Reference Wikipedia
Note: Baud should not be confused with data rate in "bits per second" (or bytes per second, etc.). Each symbol transmitted can carry one or more bits (for example, 8 bits in 256-QAM modulation) of information. When each symbol is binary it carries just one bit, so baud and bit rate are equal. This is a cheap, simple encoding. However, it's common to make better use of channel bandwidth by encoding multiple bits per symbol.
Yes but your users aren't taking the laptops out into the slum infested inner cities and leaving them inside locked vehicles while they attempt to locate people to ask questions. If the users are just taking their laptops to and from work, and maybe to their children's soccer game, then your loss rate is expected. For the Census conditions are vastly less secure, so their loss rate may be reasonable.
Usually when someone says defending the borders they mean defending against a foreign power, ie. an army. Are you saying that keeping Joe the busboy out of the country falls in the same category as repelling the British from America in 1812?
I really doubt that the foreigners are that much of a threat, and honestly you seem extremely xenophobic. If you can provide some objective evidence that proves that overall, the influx of foreigners is sufficiently detrimental that it is worthwhile to curtail their freedom (and in some sense american's freedom due to the variety of reciprocal immigration agreements we have entered into) by creating a huge bureaucracy to regulate this.
'Instriial espionnage' is not needed. The MIPS instruction set and architecture are standard reading for anybody in the field. They're not saying they copied the layout or instruction set, it's more like the API.
All processors have a language they understand, a sequence of bits that have an arbitrary meaning to them. And these are usually published far and wide, so that people can write compilers and operating systems and assemblers for this processor. MIPS in particular is very popular to study because the simple structure makes it possible for teachers to make creating a VHDL or Verilog implementation of a simplified MIPS instruction set into a half-semester project.
In fact I doubt their implementation is anywhere near the same caliber as the Pentium III implementation, even if they claim the same speed. What probably happened is they have access to more modern, smaller fabrication methods so they can cram more transistors into the pipeline. And even soft IP cores in FPGAs can hit 200MHz, so a well designed core could probably hit 500 MHz in an ASIC.
That being said, creating a full super-scalar CPU implementing even 95% of the MIPS 64-4 version of the instruction set in silicon is difficult. I was not familiar with this specification, but a quick search on google reveals that
The Mips R10000 is a dynamic superscalar microprocessor that implements the 64-bit Mips-4 Instruction Set Architecture. It fetches and decodes four instructions per cycle and dynamically issues them to five fully pipelined low-latency execution units. Instructions can be fetched and executed speculatively beyond branches. Instructions graduate in order upon completion. Although instructions execute out of order, the processor still provides sequential memory consistency and precise exception handling.The R10000 is designed for high performance, even in large real-world applications which have poor memory locality. With speculative execution, it calculates memory addresses and initiates cache refills early. Its hierarchical nonblocking memory system helps hide memory latency with two levels of set-associative, write-back caches.
This is a really beefy processor that was probably state of the art in its time. Of course that was over 10 years agos. Is it really so suprising China is only 10 years behind in chip design? After all aren't most of the chip fabrication facilities in east Asia? I'd imagine there would be quite a few people who after a few years decide they want to be on the other end of the process.
No, protocol based traffic shaping is already allowed and in use (and that is fair).
The end of net neutrality means that if you sign for VOIP service with company A, and that company doesn't pay YOUR ISP's (extortion) fee, your ISP will (at least be able to) lower your traffic quality (possibly to such a degree it is no longer functional). And maybe your ISP offers a competing VOIP service. Since they don't have to pay themselves this fee, they have an unfair market advantage (and they could set the fee to whatever they want)...
Also, it could completely disolve the peer-to-peer nature of the internet. I'm not talking about file sharing. If person A wanted to have a video-conference or whatever with person B, in order to ensure decent service, A would need to payoff B's ISP and B would need to payoff A's ISP. This sort of prior arrangement isn't very feasible in a network of peers...
In the bible quote, aren't you being a bit presumptuous in assuming the injuries were to the child, and not to the woman? In that day and age childbirth was one of the major causes of death in women, and premature childbirth was even higher. I'm just not sure how you can justify that this quote is applicable...
If we can just pull out bible quotes to support our points without careful consideration as to what they really meant in context, then here's part of one from Lamentations: "It's better to die a babe unborn..."
I'm not sure why you think it's xml...The only time you even see an > or < is if you click buttons and that just means "type something here".
I never use the mouse to hunt and click to input formulas. The syntax is very simple once you get used to it, and is very similar to latex.
It's relatively easy to type something like
{-b +- sqrt{b^2 - 4 a c} } over {2 a}
(and arguably more intuitive than the latex equivalent:
\frac{-b \pm \sqrt{b^2 - 4 a c}}{2 a}
)
And I've had more luck getting the summation ranges looking pretty in OO than in latex for equations like:
{sum from{k=-infinity} to{infinity} { a sub k e^{j w sub T k}}} over {sum from{i=-infinity} to{infinity} {a sub k}^2}
But even if you do your research, finding a wireless device with good Linux drivers is a crapshoot. Manufacturers like to change the chipset without changing the product number or revision number. When I was shopping for a pci wireless card I read reviews and checked the lists and decided to buy a card with the prism54 drivers (which are the most supported). I decided on a Hawking HWP54G which supposedly uses the prism54 chip, although there were reports newer ones used the broadcom chip (which is probably the second most supported). When it arrived I plugged it in and found to my delight it had a crappy Texas Instruments ACX 111. Fortunately some kind soul had hacked out a partially working driver, so it is moderately functional -- I eventually got it to pass traffic after hacking my init scripts, but dhcp always times out, it dumps tons of error messages in the kernel log, the driver seems to randomly stop functioning sometimes and I have to unload and reload the driver for it to start working again, additionally my system is noticeably less stable (but that appears to be improving with the latest release).
Of course this is more or less true with every piece of hardware, since hardware vendors like to switch suppliers to reduce costs and abstract the difference in software. It is more of a problem for wireless where Linux drivers are more sparse...
A world without religion. A world where lazy people still believe whatever tells them. A world where leaders can still start a war arbitrarily, citing security or nationalism. A world where people still starve while some are still hungry. A world where people still feel restricted in their actions over concern for what society thinks.
Hello new world, same as the old world....
Your problem is not with religion. Your problem is with hypocracy, laziness, greed and many other evils. Religion by itself is neither good nor evil.
But Triple-ROT52 only works with a small subset of 7-bit ascii plaintext. I think they're using either Quadruple-ROT64 (supports 8bit ascii) or Octuple-ROT536870912 (for proper unicode support). Either way it's kind of silly to use a data-dependent algorithm here...
"The area where your average linux distro needs to improve is disk space. There is no excuse for the default desktop installation to take over 500mb. A great deal of this is caused by programmer never reinventing wheels in OSS. This means that every program has dependancies for a half dozen third party programs and libraries instead of incorporating just the functionality they need for their app. Those half dozen programs and libraries have followed the same bloat yielding philosophy and have their own dependancies; those in turn have their own dependancies and so on."
So you're seriously saying that in order to improve disk space we need create many copies of the same functionality? This is very unintuitive....And I'm not sure how this could possibly be so, except if the libraries were huge and only selected to offer 1 or 2 functions.. But it seems to me, that if there are many dependencies (as you claim), it would suggest functionality was broken up into small pieces, and the designers merely selected the pieces they required. This is in contrast to large frameworks like.Net or whatever where you have to have the entire framework. So which is it? Is the functionality not divided finely enough (note that dividing it further increases start-up time due to linking more libraries) or is it divided too finely (note that this will increase disk usage due to waste).
Just because a 1kB program requires 12-15 100kB libraries (and maybe the requisite 3-5MB gtk/or whatever toolkit) doesn't mean it wastes space. Assuming the functionality is sufficiently finely divided and library overhead is negligible, nothing is lost this way in the worst case. In the best case it results in (primarily) more efficient RAM usage and also more efficient harddrive usage.
This is outside my expertise, but this is what I think gp means: Your equation is correct, but incomplete. It neglects the relationship between voltage and frequency. The frequency of a synchrous sequential circuit is limited by amount of time it takes for the signal to propogate through the worst-case path. The delay is caused by the combination of the channel resistance of a stage and the gate capacitance of the following stage (plus parasitics and non-linearities). This forms a basic RC circuit so the voltage will follow an exponential curve. The signal will only move to the next stage when the voltage passes the threshold voltage for the next stage (plus an additional switching delay). Increasing the voltage will scale up the curve but not change the threshold. To a first order approximation this makes voltage proportional to frequency. If you then plug this in for voltage in your equation you'll get the gp's result.
Rather than insulting a poster who was politely suggesting something they read (which if you wanted you could google for), why don't you think for yourself about under what conditions what they're saying makes sense?
You posted that states should decide what murder is, but then went on to say that this is okay because if they did something wrong the federal government would decide it was unconstitutional. If the decision were purely up to the state, there would be no federal or constitutional oversight. States could do whatever they wanted. It makes no sense to say that states should have absolute authority in something as long as it agrees with what the federal government or Constitution says.
Our founding fathers tried to create weak decentralized government under the Articles of Confederation, and failed. Our nation should not try to completely define sensitive issues that are open to interpretation, but instead should provide a rough framework to restrain and support more localized frameworks (Similar to the way all the rooms in a house have common framing and support, but may have different walls, carpeting, sizes and shapes).
I agree, Linux sure is easy these days.
Why just last week I was upgrading Xorg on my Gentoo box, and only had to recompile my entire system (including kernel) and copy/usr/share/fonts/misc from a working X installation, and edit some config files.
Next week I may just see about upgrading gcc from 3.3 to 3.4.
Thank heavens I live in a modern era.
(Of course I actually like gentoo, and for most things it is simple to maintain, but some things are not quite...)
Actually not all encrypted traffic streams are equal. All encryption does is hide the contents of the message from observation. Even though every packet looks the same, the stream characteristics makes it vulnerable to statistical classification.
By stream characteristics I mean timing of packets, bandwidth, endpoints, directional parity, and things like that.
Examples: If a stream has real-time characteristics (ie. packets are roughly equally spaced in time, use more realtime protocol like UDP, etc), is bidirectional and moderate bandwidth demands it is probably VoIP. If it's only unidirectional it is almost certainly streaming audio. If it has high bandwidth instead, maybe it's video conferencing. If it's bandwidth comes in bursts of requests, it's probably HTTP, FTP, IMAP, or something like that. and If it takes up as much bandwidth as is a available in any direction, it's bittorrent.
Additionally VoIP has certain servers that connect it with the POTS network. So its trivial to detect when one of these are used.
There are ways to get around this (onion routing, etc) but they are generally seen as impractical for real-time streams.
Wizards complicate the issue. A person beginning to program has difficulty understanding the basic syntax that they themselves are using, much less trying to understand the advanced code the wizard creates so they can figure out where it went wrong. If you want to learn, start simple.
You don't learn to fly using the autopilot of a 747.
The submitter didn't even get to the second page. This technology was originally designed to dampen mechanical vibration in ships like the submitter said, but the Navy wasn't interested (probably because while that would be useful in a cold war, it's not too helpful in the war on terror).
What they're doing now, is using to cushion seats. The article claims some of the speedier boats they use to put marines ashore can pull 8Gs hopping over waves. So they use this technology along with wave height sensors and a fast processor to reduce it to 2-3Gs. Although this sounds new, it's basically the same technique as those active noise cancellation headphones only for lower frequencies and higher amplitudes.
Then the company is trying to expand by putting this into other vehicles like humvees and helicopters. (why in a helicopter?)
Slashdot Mirror
Current systems only have a name in the database.
What makes you think that the new system will have pictures, a name and a text description when the current system only has one of them?
Modern cellphones adjust the transmit power based on the Signal-Noise Ration. If there are a lot of cellphones around, then the noise is high, so the phone has to increase the transmit power (reducing battery life). If the cell phone is able to find a frequency with very little noise, it can transmit with very little power, and still have the same quality. If a software radio is used, the number of possible frequencies becomes much larger.
The gains from using this method should more than offset the losses from the slightly less efficient analog stages and higher-speed digital processors and converters.
It'd be proof that man can't spell
Or write a Spamming IE plugin
Reference Wikipedia
Note: Baud should not be confused with data rate in "bits per second" (or bytes per second, etc.). Each symbol transmitted can carry one or more bits (for example, 8 bits in 256-QAM modulation) of information. When each symbol is binary it carries just one bit, so baud and bit rate are equal. This is a cheap, simple encoding. However, it's common to make better use of channel bandwidth by encoding multiple bits per symbol.
"If you reboot your machine remotely, and forget that the last option selected has no network support, you have no way to access the machine."
Question:
How do you reboot the machine remotely from a configuration that has no network support?
Yes but your users aren't taking the laptops out into the slum infested inner cities and leaving them inside locked vehicles while they attempt to locate people to ask questions. If the users are just taking their laptops to and from work, and maybe to their children's soccer game, then your loss rate is expected. For the Census conditions are vastly less secure, so their loss rate may be reasonable.
Usually when someone says defending the borders they mean defending against a foreign power, ie. an army.
Are you saying that keeping Joe the busboy out of the country falls in the same category as repelling the British from America in 1812?
I really doubt that the foreigners are that much of a threat, and honestly you seem extremely xenophobic. If you can provide some objective evidence that proves that overall, the influx of foreigners is sufficiently detrimental that it is worthwhile to curtail their freedom (and in some sense american's freedom due to the variety of reciprocal immigration agreements we have entered into) by creating a huge bureaucracy to regulate this.
All processors have a language they understand, a sequence of bits that have an arbitrary meaning to them. And these are usually published far and wide, so that people can write compilers and operating systems and assemblers for this processor. MIPS in particular is very popular to study because the simple structure makes it possible for teachers to make creating a VHDL or Verilog implementation of a simplified MIPS instruction set into a half-semester project.
In fact I doubt their implementation is anywhere near the same caliber as the Pentium III implementation, even if they claim the same speed. What probably happened is they have access to more modern, smaller fabrication methods so they can cram more transistors into the pipeline. And even soft IP cores in FPGAs can hit 200MHz, so a well designed core could probably hit 500 MHz in an ASIC.
That being said, creating a full super-scalar CPU implementing even 95% of the MIPS 64-4 version of the instruction set in silicon is difficult. I was not familiar with this specification, but a quick search on google reveals that This is a really beefy processor that was probably state of the art in its time. Of course that was over 10 years agos. Is it really so suprising China is only 10 years behind in chip design? After all aren't most of the chip fabrication facilities in east Asia? I'd imagine there would be quite a few people who after a few years decide they want to be on the other end of the process.
So what, after they shoot you a few thousand times you can crack the key?
No, protocol based traffic shaping is already allowed and in use (and that is fair).
The end of net neutrality means that if you sign for VOIP service with company A, and that company doesn't pay YOUR ISP's (extortion) fee, your ISP will (at least be able to) lower your traffic quality (possibly to such a degree it is no longer functional). And maybe your ISP offers a competing VOIP service. Since they don't have to pay themselves this fee, they have an unfair market advantage (and they could set the fee to whatever they want)...
Also, it could completely disolve the peer-to-peer nature of the internet. I'm not talking about file sharing. If person A wanted to have a video-conference or whatever with person B, in order to ensure decent service, A would need to payoff B's ISP and B would need to payoff A's ISP.
This sort of prior arrangement isn't very feasible in a network of peers...
In the bible quote, aren't you being a bit presumptuous in assuming the injuries were to the child, and not to the woman?
In that day and age childbirth was one of the major causes of death in women, and premature childbirth was even higher.
I'm just not sure how you can justify that this quote is applicable...
If we can just pull out bible quotes to support our points without careful consideration as to what they really meant in context, then here's part of one from Lamentations:
"It's better to die a babe unborn..."
I'm not sure why you think it's xml...The only time you even see an > or < is if you click buttons and that just means "type something here".
I never use the mouse to hunt and click to input formulas. The syntax is very simple once you get used to it, and is very similar to latex.
It's relatively easy to type something like
(and arguably more intuitive than the latex equivalent: )
And I've had more luck getting the summation ranges looking pretty in OO than in latex for equations like:
But even if you do your research, finding a wireless device with good Linux drivers is a crapshoot.
Manufacturers like to change the chipset without changing the product number or revision number.
When I was shopping for a pci wireless card I read reviews and checked the lists and decided to buy a card with the prism54 drivers (which are the most supported). I decided on a Hawking HWP54G which supposedly uses the prism54 chip, although there were reports newer ones used the broadcom chip (which is probably the second most supported).
When it arrived I plugged it in and found to my delight it had a crappy Texas Instruments ACX 111.
Fortunately some kind soul had hacked out a partially working driver, so it is moderately functional -- I eventually got it to pass traffic after hacking my init scripts, but dhcp always times out, it dumps tons of error messages in the kernel log, the driver seems to randomly stop functioning sometimes and I have to unload and reload the driver for it to start working again, additionally my system is noticeably less stable (but that appears to be improving with the latest release).
Of course this is more or less true with every piece of hardware, since hardware vendors like to switch suppliers to reduce costs and abstract the difference in software. It is more of a problem for wireless where Linux drivers are more sparse...
A world without religion.
A world where lazy people still believe whatever tells them.
A world where leaders can still start a war arbitrarily, citing security or nationalism.
A world where people still starve while some are still hungry.
A world where people still feel restricted in their actions over concern for what society thinks.
Hello new world, same as the old world....
Your problem is not with religion.
Your problem is with hypocracy, laziness, greed and many other evils.
Religion by itself is neither good nor evil.
But Triple-ROT52 only works with a small subset of 7-bit ascii plaintext.
I think they're using either Quadruple-ROT64 (supports 8bit ascii) or Octuple-ROT536870912 (for proper unicode support).
Either way it's kind of silly to use a data-dependent algorithm here...
It was pretty bad.
"The area where your average linux distro needs to improve is disk space. There is no excuse for the default desktop installation to take over 500mb. A great deal of this is caused by programmer never reinventing wheels in OSS. This means that every program has dependancies for a half dozen third party programs and libraries instead of incorporating just the functionality they need for their app. Those half dozen programs and libraries have followed the same bloat yielding philosophy and have their own dependancies; those in turn have their own dependancies and so on."
.Net or whatever where you have to have the entire framework.
So you're seriously saying that in order to improve disk space we need create many copies of the same functionality?
This is very unintuitive....And I'm not sure how this could possibly be so, except if the libraries were huge and only selected to offer 1 or 2 functions..
But it seems to me, that if there are many dependencies (as you claim), it would suggest functionality was broken up into small pieces, and the designers merely selected the pieces they required. This is in contrast to large frameworks like
So which is it? Is the functionality not divided finely enough (note that dividing it further increases start-up time due to linking more libraries) or is it divided too finely (note that this will increase disk usage due to waste).
Just because a 1kB program requires 12-15 100kB libraries (and maybe the requisite 3-5MB gtk/or whatever toolkit) doesn't mean it wastes space.
Assuming the functionality is sufficiently finely divided and library overhead is negligible, nothing is lost this way in the worst case. In the best case it results in (primarily) more efficient RAM usage and also more efficient harddrive usage.
This is outside my expertise, but this is what I think gp means:
Your equation is correct, but incomplete. It neglects the relationship between voltage and frequency.
The frequency of a synchrous sequential circuit is limited by amount of time it takes for the signal to propogate through the worst-case path.
The delay is caused by the combination of the channel resistance of a stage and the gate capacitance of the following stage (plus parasitics and non-linearities). This forms a basic RC circuit so the voltage will follow an exponential curve. The signal will only move to the next stage when the voltage passes the threshold voltage for the next stage (plus an additional switching delay). Increasing the voltage will scale up the curve but not change the threshold. To a first order approximation this makes voltage proportional to frequency.
If you then plug this in for voltage in your equation you'll get the gp's result.
Rather than insulting a poster who was politely suggesting something they read (which if you wanted you could google for), why don't you think for yourself about under what conditions what they're saying makes sense?
You posted that states should decide what murder is, but then went on to say that this is okay because if they did something wrong the federal government would decide it was unconstitutional.
If the decision were purely up to the state, there would be no federal or constitutional oversight. States could do whatever they wanted.
It makes no sense to say that states should have absolute authority in something as long as it agrees with what the federal government or Constitution says.
Our founding fathers tried to create weak decentralized government under the Articles of Confederation, and failed.
Our nation should not try to completely define sensitive issues that are open to interpretation, but instead should provide a rough framework to restrain and support more localized frameworks (Similar to the way all the rooms in a house have common framing and support, but may have different walls, carpeting, sizes and shapes).
I agree, Linux sure is easy these days. /usr/share/fonts/misc from a working X installation, and edit some config files.
Why just last week I was upgrading Xorg on my Gentoo box, and only had to recompile my entire system (including kernel) and copy
Next week I may just see about upgrading gcc from 3.3 to 3.4.
Thank heavens I live in a modern era.
(Of course I actually like gentoo, and for most things it is simple to maintain, but some things are not quite...)
Actually not all encrypted traffic streams are equal. All encryption does is hide the contents of the message from observation. Even though every packet looks the same, the stream characteristics makes it vulnerable to statistical classification.
By stream characteristics I mean timing of packets, bandwidth, endpoints, directional parity, and things like that.
Examples:
If a stream has real-time characteristics (ie. packets are roughly equally spaced in time, use more realtime protocol like UDP, etc), is bidirectional and moderate bandwidth demands it is probably VoIP. If it's only unidirectional it is almost certainly streaming audio.
If it has high bandwidth instead, maybe it's video conferencing.
If it's bandwidth comes in bursts of requests, it's probably HTTP, FTP, IMAP, or something like that.
and If it takes up as much bandwidth as is a available in any direction, it's bittorrent.
Additionally VoIP has certain servers that connect it with the POTS network. So its trivial to detect when one of these are used.
There are ways to get around this (onion routing, etc) but they are generally seen as impractical for real-time streams.
"[T]he concept or cache-misses can be taught regardless of language."
Please tell me how.
And while you're at it, could you teach me the difference between real-mode and protected-mode in Java?
Thanks
Wizards complicate the issue.
A person beginning to program has difficulty understanding the basic syntax that they themselves are using, much less trying to understand the advanced code the wizard creates so they can figure out where it went wrong.
If you want to learn, start simple.
You don't learn to fly using the autopilot of a 747.
The submitter didn't even get to the second page.
This technology was originally designed to dampen mechanical vibration in ships like the submitter said, but the Navy wasn't interested (probably because while that would be useful in a cold war, it's not too helpful in the war on terror).
What they're doing now, is using to cushion seats. The article claims some of the speedier boats they use to put marines ashore can pull 8Gs hopping over waves. So they use this technology along with wave height sensors and a fast processor to reduce it to 2-3Gs. Although this sounds new, it's basically the same technique as those active noise cancellation headphones only for lower frequencies and higher amplitudes.
Then the company is trying to expand by putting this into other vehicles like humvees and helicopters. (why in a helicopter?)