If you calculate the data twice, that gets you from a.2% error rate to at.04% rate. At a billion operations per second, that's still 400,000 errors per second.
What DNA Computing needs is proper error correction codes, and this is mostly an open problem. Here's a quick Google Search on that topic.
Also, maybe that computer is fast and tiny, it won't handle a lot of data. Classical NP-complete problems maybe aren't time complex anymore because of massive parallelism, but you get a 'weight' complexity that often severely limits the problem size.
And we shouldn't forget that extracting the output from the test tube is currently a very slow process.
Aside from the fact that it's most likely a hoax, I'm not entirely sure that it's a big step forward for Afghans to be able to receive Baywatch, Survivor, and Temptation Island.
Think about how Linux first came into existence. Just for fun. People just like hacking and playing around with stuff, there doesn't necessarily have to be a good reason for it at first. Most important scientific discoveries came out of originally 'useless' research.
There is no comparison with Napster. Napster was blatantly illegal and deserved to be shut down. It was just a warez site for music.
While you are right, I think Yu Suzuki is making a good point using Napster as an analogy. Start in the Underground with a cool project, get famous, then become absorbed and destroyed by the Mainstream and Mainstream business lobbyists. Luckily enough, though, Linux is very different from things like Napster, and therefore I'm very sure it won't grow up to be destroyed.
I think the point of doing the PhD is not to go get a normal job after that, but rather stay in the research line.
But anyway, I think it is important to note that the goal to aim for when getting a CS degree is not just being a programmer, but much rather being an Architect. Sure you can code. But I think that people like Linus also have lots of fun because they don't only get to implement stuff, but they get to design and create.
There's so much more than just coding and implementing stuff. Just look around yourself, there's thousands of possibilities.
Very true, if you look at all the subnotebooks that featured Transmeta chips. Low power requirements are just useless if the device is so small that the battery is crippled. And for those kind of devices, there are those other processors you mentioned. I would've preferred seeing a Crusoe in more regular-sized hardware, just to finally get a drop-down in temperature and longer uptime while not connected to other power sources. But instead, regular notebooks come with fat Gigahertz processors that I can't really see a good use for - at least I plan to work and code with my notebook, not encode or watch DivX stuff and play state-of-the-art 3D games.
Daaamn. And I was still hoping that maybe someone would finally build up a nice notebook, regular size, regular battery, Crusoe, that wouldn't burn my legs while working, and run for more than 3 hours. Well, I guess not... looks like it'll be an iBook for me then.
I think this is gonna be rather easy. They're gonna try again and again to keep us from using music as we want to, and they will fail each and every time. Because we will always find a way around their obstacles. And one day, they're just gonna give up. It may take a while, but they will, after they notice that throwing tons of money into coyright-protection mechanisms won't pay off.
Most likely, steg data is detectable in images with areas of similar color or continuous gradients. But will it be detectable in, say, a.wav containing white noise, or an image filled with random data? Of course, mailing useless images or mp3s with noise around would be a little suspicious.
Even if GPS were to work in a place like that, I think that firstly the accuracy is not high enough to find out exactly under which piece of debris a person is hidden, and secondly GPS will not tell you how deep a person lies.
Just in case anybody wonders...
on
KDE 2.2 Tagged
·
· Score: 2, Informative
If anybody wonders what Fitt's law is, here ya go.
I'm kinda surprised that, when prompted about similarities to Communism, RMS starts talking about the Soviet system. I think this is a rather common misconception.
Communism itself has exactly nothing to do with the Soviet system. The Soviets have never run a true Communistic system. When people refer to Communism they often erroneously refer to the Soviet Union and think about the political system.
Communism is not a political system but rather an economic philosophy that, in my opinion, has quite a lot of things in common with the Free Software philosophy, in that it allows people to share their goods freely. It is not an alternative to, say, monarchies or democracy, because it is no political system. It is rather an alternative to Capitalism.
Free Software definitely has nothing to do with the Soviet (political) system. But neither does Communism.
I haven't spent more than $20 on Linux distros. I bought my first Linux three years ago, and half a year later a Mandrake CD. Since then the 10 MBit line in my college dorm does it all... rpmfind.net or, for my other box, apt-get are better than any CD.
...which leaves one imho pretty interesting question:
Assuming we had a proper secure, privacy-respecting, standardized, well-spread and easy-to-use electronic cash system, how would internet services change, how would prices develop, how would people behave? I for one would be happy to see such a system and would most likely be willing to pay for certain services if the prices were reasonable.
Maybe these guys should rethink their database structure... six Terabytes for just nine million users means more than half a meg of data per person... I wonder what all the stuff is they have stored...
Well it's not always stupid, but maybe unnecessary. Basically all you need is a secure (4096-Bit) PGP Key with a long and very secure password. Then make a list of all other less important passwords and gpg-encrypt them. So all you need to remember is that one password. In case you're a little more paranoid, get the int'l kernel patch and create an AES or RC6 encrypted loopmount, and put the gpg-encrypted password list on that. Should be safe enough for most cases.
Easy. here's a recent example I had. I noticed that I hadn't installed curl on a debian box. So here ya go: apt-get install curl. Done. Fresh from the net. Questions?
> (I take slashdot's cookie. All others, I accept
> and discard when shutting down the browser.)
That's exactly what I want to see in a browser. Many shopping sites need cookies to keep track of the shopping cart, which is fine with me. And I don't want to keep turning cookies on and off or manually accepting/rejecting them just to be able to use a shopping cart. My current solution is starting netscape/mozilla via a script that deletes all stored cookies except the ones I want to keep (/. etc). That way, cookie-dependent web apps will work, but cookies will disappear the next browsing session.
Hehehe... it's back in effect now... see also here, hope the fish helps.
If you calculate the data twice, that gets you from a .2% error rate to at .04% rate. At a billion operations per second, that's still 400,000 errors per second.
What DNA Computing needs is proper error correction codes, and this is mostly an open problem. Here's a quick Google Search on that topic.
Also, maybe that computer is fast and tiny, it won't handle a lot of data. Classical NP-complete problems maybe aren't time complex anymore because of massive parallelism, but you get a 'weight' complexity that often severely limits the problem size.
And we shouldn't forget that extracting the output from the test tube is currently a very slow process.
Aside from the fact that it's most likely a hoax, I'm not entirely sure that it's a big step forward for Afghans to be able to receive Baywatch, Survivor, and Temptation Island.
I actually feel kind of sorry for them.
Think about how Linux first came into existence. Just for fun. People just like hacking and playing around with stuff, there doesn't necessarily have to be a good reason for it at first. Most important scientific discoveries came out of originally 'useless' research.
There is no comparison with Napster. Napster was blatantly illegal and deserved to be shut down. It was just a warez site for music.
While you are right, I think Yu Suzuki is making a good point using Napster as an analogy. Start in the Underground with a cool project, get famous, then become absorbed and destroyed by the Mainstream and Mainstream business lobbyists. Luckily enough, though, Linux is very different from things like Napster, and therefore I'm very sure it won't grow up to be destroyed.
Jimmy Eat World? Absolutely no trouble with that one. cdda2wav worked perfectly on my TEAC CD-R58S.
I think the point of doing the PhD is not to go get a normal job after that, but rather stay in the research line.
But anyway, I think it is important to note that the goal to aim for when getting a CS degree is not just being a programmer, but much rather being an Architect. Sure you can code. But I think that people like Linus also have lots of fun because they don't only get to implement stuff, but they get to design and create.
There's so much more than just coding and implementing stuff. Just look around yourself, there's thousands of possibilities.
Very true, if you look at all the subnotebooks that featured Transmeta chips. Low power requirements are just useless if the device is so small that the battery is crippled. And for those kind of devices, there are those other processors you mentioned. I would've preferred seeing a Crusoe in more regular-sized hardware, just to finally get a drop-down in temperature and longer uptime while not connected to other power sources. But instead, regular notebooks come with fat Gigahertz processors that I can't really see a good use for - at least I plan to work and code with my notebook, not encode or watch DivX stuff and play state-of-the-art 3D games.
Daaamn. And I was still hoping that maybe someone would finally build up a nice notebook, regular size, regular battery, Crusoe, that wouldn't burn my legs while working, and run for more than 3 hours. Well, I guess not... looks like it'll be an iBook for me then.
I think this is gonna be rather easy. They're gonna try again and again to keep us from using music as we want to, and they will fail each and every time. Because we will always find a way around their obstacles. And one day, they're just gonna give up. It may take a while, but they will, after they notice that throwing tons of money into coyright-protection mechanisms won't pay off.
[...] and no one has tried to benefit by it.
Except for the US weapons industry, of course... private firearms, gas masks...
Just to remind you of this project...
OpenMusic
Most likely, steg data is detectable in images with areas of similar color or continuous gradients. But will it be detectable in, say, a .wav containing white noise, or an image filled with random data? Of course, mailing useless images or mp3s with noise around would be a little suspicious.
Even if GPS were to work in a place like that, I think that firstly the accuracy is not high enough to find out exactly under which piece of debris a person is hidden, and secondly GPS will not tell you how deep a person lies.
If anybody wonders what Fitt's law is, here ya go.
I'm kinda surprised that, when prompted about similarities to Communism, RMS starts talking about the Soviet system. I think this is a rather common misconception.
Communism itself has exactly nothing to do with the Soviet system. The Soviets have never run a true Communistic system. When people refer to Communism they often erroneously refer to the Soviet Union and think about the political system.
Communism is not a political system but rather an economic philosophy that, in my opinion, has quite a lot of things in common with the Free Software philosophy, in that it allows people to share their goods freely. It is not an alternative to, say, monarchies or democracy, because it is no political system. It is rather an alternative to Capitalism.
Free Software definitely has nothing to do with the Soviet (political) system. But neither does Communism.
I haven't spent more than $20 on Linux distros. I bought my first Linux three years ago, and half a year later a Mandrake CD. Since then the 10 MBit line in my college dorm does it all... rpmfind.net or, for my other box, apt-get are better than any CD.
That's what MS was thinking, right? .NET C#" (pronounce: net cash)
"We love
...which leaves one imho pretty interesting question:
Assuming we had a proper secure, privacy-respecting, standardized, well-spread and easy-to-use electronic cash system, how would internet services change, how would prices develop, how would people behave? I for one would be happy to see such a system and would most likely be willing to pay for certain services if the prices were reasonable.
Or:
curl --user-agent "Download Manager" http://svmsftwxp.conxion.com/download/wxp_pro_rc1. iso > wxp_pro_rc1.iso
Maybe these guys should rethink their database structure... six Terabytes for just nine million users means more than half a meg of data per person... I wonder what all the stuff is they have stored...
Well, actually you should consider that all games using 3D are practically vector based games ;-)
Well it's not always stupid, but maybe unnecessary. Basically all you need is a secure (4096-Bit) PGP Key with a long and very secure password. Then make a list of all other less important passwords and gpg-encrypt them. So all you need to remember is that one password. In case you're a little more paranoid, get the int'l kernel patch and create an AES or RC6 encrypted loopmount, and put the gpg-encrypted password list on that. Should be safe enough for most cases.
Easy. here's a recent example I had. I noticed that I hadn't installed curl on a debian box. So here ya go: apt-get install curl. Done. Fresh from the net. Questions?
> (I take slashdot's cookie. All others, I accept
> and discard when shutting down the browser.)
That's exactly what I want to see in a browser. Many shopping sites need cookies to keep track of the shopping cart, which is fine with me. And I don't want to keep turning cookies on and off or manually accepting/rejecting them just to be able to use a shopping cart. My current solution is starting netscape/mozilla via a script that deletes all stored cookies except the ones I want to keep (/. etc). That way, cookie-dependent web apps will work, but cookies will disappear the next browsing session.