I have been toying with the following idea for some time now...
Form a non-profit organization. Everyone who doesn't want to receive SPAM will send this organization an authorization to act on his/her behalf.
Whenever you receive SPAM, you forward it to this organization (headers and all). No big deal. A few seconds of work.
This organization then traces the spammer (the spammer has to leave a fingerprint, as they want to sell something). This organization then files a lawsuit against this spammer, asking for 10c for each member that was spammed. This money is used to pay the lawyers of the organization.
I am hoping that the spammers will find they can no longer send millions of ads for a trivial cost. They now have to pay for lawyers, stand in court, and potentially, pay thousands of dollars for their abuse. I am hoping that this, high, cost will deter spammers, and can, potentially, kill the entire spam phenomena altogether.
The only problem is that I am not a US citizen (or resident), and I cannot setup such an organization.
Isn't it a bit too early to say the MS era is over? After all, there is a long road ahead, even for MS.
What is interesting to me, however, is what happens if MS does lose. Furthermore, what happens if it goes bankrupt becuase of punishing compensation lawsuites from dozens, if not hundreds, of allegably offended parties?
First of all, an explanation with a disclaimer. Both me and Fidros are private people working for a company that was mentioned here, and decided (totally uncoordinated between us) to answer. This is not an answer from GTek, this is an answer from the actual people. Neither of us comes from marketting. In fact, we are both developers, and have been personally involved with some of the projects mentioned here, and others that were not.
To put things further into prespective, GTek is a software company, with no more than 15 full time software developers. We do not have a PR machine. In fact, the total number of PR plus marketting personnel we employ is zero.
I should further point out that noone is sueing anyone. You may notice that my reply contained no mention of it, and Fidros' reply had a smily after mentioning it. I should also note that both "Fidros" and "Sun" are our standard/. nicks. We are regular slashdot readers, and this is the way we came across this story here.
As for the facts. To the best of my knowledge, there are very few companies that can do both hardware and software. I.B.M. is rumoured to be one, I will possibly include Intel as well. The rest of us have to choose. Companies like PB (and most other OEMs as well) need a company to provide the software.
GTek is such a company. We provide software products tailor made to another company's demands, and we don't even place our logo on these applications. This means that PB got to decide exactly what features were in, and how the product would look. To my opinion, this also means that they should have participated in the development process. At the very least, you could have expected proper feedback about what products are generating Technical Support calls, and what the common problems were.
I will not go into details. For one thing, as I have said before, I do not represent GTek. Another is that I do not have intemate knowledge about everything, I would not like to slander anyone by mistake. Suffice it to say that nasty politics were a major part of GTek's dealings with PB. I personally tried to avoid such matters, but it was not always possible (for example, when being greated by one VP with a big, fatherly, friendly "Hello, why are you not meeting your schedule?", when I have no project in common with said VP, and in general, am not behind on any schedule).
Like any other company with more then one product, some of GTek's products were better. Some were worst. Certainly, everyone has the right to an opinion about them. All in all, I think GTek's role in PackardBell's history did a lot more to help it survive as long as it did than to pull it down.
First, as can be deducted from my home URL, I work for GTek. I must say that constructive criticism has always been a good thing in people.
In case anyone is interested, we have been producing many applications that were shipped with PB computers. One point I cannot stress enough, though. We are not the company responsible for "PackardBell Navigator". We have nothing to do with that company, or with the product.
Throughout our buisness with PB we have had to deal with internal PB politics. As a result, we would hear about angry (sometimes abusive) letters released by various PB devisions about our products, but repeated requests for feedback about the actual problems were almost always denied.
The way I see it - If you want total control over product customizations, you have to participate in the QA. At the very least - tell your provider what the problems are.
One last note - GTek Technologies Ltd is an Israeli company. You can find our main page here over a very slow connection. This web site is mirrored, at a much faster rate and in the US, here. We have nothing to do with an American company called "GTek Inc." (you can find their site here.
I understand it covers the lempel-ziv welsh patent, but is that the only variant covered? What about Lempel-Ziv huffman (LZH)? What exactly is the welsh varient?
Actually, the "implement on cheap hardware" requirment is not an AES requirment at all.
The "requirment" was introduced by Bruce and his friends in his twofish proposal. I have not found a trace of this requirment in the original AES requirments.
On page 14 of the TwoFish proposal, titled "Language, Compiler, and Processor Choice", they say "It is clear that the Borland C 5.0 compiler chosen as the standard AES reference is not the best optimizing compiler."...
Optimizing for small hardware was introduced by Bruce and friends, perhaps to distinguish their submission from other submissions?
The way I understand this, Euler Phi is not the limiting factor either. The limiting factor is the ability to perform DRoot (descreat root).
In order to perform DRoot, you need Euler's Phi, and in order to get that, you need to factor the public key. This is, of course, unless someone finds a better way.
The point is, that someone may find a way to do DRoot, and bypass the factoring, as well as Euler's Phi, problem.
I think we can conclude that factoring complexity >= finding Phi >= finding DRoot.
As for Diffie-Helman, it is based on the difficulty of DLog. The base modulo for DH is a prime number itself. Factoring, therefor, does not enter into it at all.
You tell me how good they are. DES was not publicly reviewed before it was released. How long did it take the rest of the world to find diffrential cryptanalisis (why IBM and the NSA knew about it when DES was invented)?
Slashdot Mirror
According to Mindcraft's tests, the bottleneck was the Kernel. In particular, the TCP/IP stack is single threaded.
Form a non-profit organization. Everyone who doesn't want to receive SPAM will send this organization an authorization to act on his/her behalf.
Whenever you receive SPAM, you forward it to this organization (headers and all). No big deal. A few seconds of work.
This organization then traces the spammer (the spammer has to leave a fingerprint, as they want to sell something). This organization then files a lawsuit against this spammer, asking for 10c for each member that was spammed. This money is used to pay the lawyers of the organization.
I am hoping that the spammers will find they can no longer send millions of ads for a trivial cost. They now have to pay for lawyers, stand in court, and potentially, pay thousands of dollars for their abuse. I am hoping that this, high, cost will deter spammers, and can, potentially, kill the entire spam phenomena altogether.
The only problem is that I am not a US citizen (or resident), and I cannot setup such an organization.
Please tell me what you think.
Isn't it a bit too early to say the MS era is over? After all, there is a long road ahead, even for MS.
What is interesting to me, however, is what happens if MS does lose. Furthermore, what happens if it goes bankrupt becuase of punishing compensation lawsuites from dozens, if not hundreds, of allegably offended parties?
First of all, an explanation with a disclaimer. Both me and Fidros are private people working for a company that was mentioned here, and decided (totally uncoordinated between us) to answer. This is not an answer from GTek, this is an answer from the actual people. Neither of us comes from marketting. In fact, we are both developers, and have been personally involved with some of the projects mentioned here, and others that were not.
To put things further into prespective, GTek is a software company, with no more than 15 full time software developers. We do not have a PR machine. In fact, the total number of PR plus marketting personnel we employ is zero.
I should further point out that noone is sueing anyone. You may notice that my reply contained no mention of it, and Fidros' reply had a smily after mentioning it. I should also note that both "Fidros" and "Sun" are our standard /. nicks. We are regular slashdot readers, and this is the way we came across this story here.
As for the facts. To the best of my knowledge, there are very few companies that can do both hardware and software. I.B.M. is rumoured to be one, I will possibly include Intel as well. The rest of us have to choose. Companies like PB (and most other OEMs as well) need a company to provide the software.
GTek is such a company. We provide software products tailor made to another company's demands, and we don't even place our logo on these applications. This means that PB got to decide exactly what features were in, and how the product would look. To my opinion, this also means that they should have participated in the development process. At the very least, you could have expected proper feedback about what products are generating Technical Support calls, and what the common problems were.
I will not go into details. For one thing, as I have said before, I do not represent GTek. Another is that I do not have intemate knowledge about everything, I would not like to slander anyone by mistake. Suffice it to say that nasty politics were a major part of GTek's dealings with PB. I personally tried to avoid such matters, but it was not always possible (for example, when being greated by one VP with a big, fatherly, friendly "Hello, why are you not meeting your schedule?", when I have no project in common with said VP, and in general, am not behind on any schedule).
Like any other company with more then one product, some of GTek's products were better. Some were worst. Certainly, everyone has the right to an opinion about them. All in all, I think GTek's role in PackardBell's history did a lot more to help it survive as long as it did than to pull it down.
Shachar Shemesh
sun-sdr@gtek.co.il
P.S.
As this thread is quickly deteriorating into flame wars, I would prefare it if further replies were sent to me via e-mail.
In case anyone is interested, we have been producing many applications that were shipped with PB computers. One point I cannot stress enough, though. We are not the company responsible for "PackardBell Navigator". We have nothing to do with that company, or with the product.
Throughout our buisness with PB we have had to deal with internal PB politics. As a result, we would hear about angry (sometimes abusive) letters released by various PB devisions about our products, but repeated requests for feedback about the actual problems were almost always denied.
The way I see it - If you want total control over product customizations, you have to participate in the QA. At the very least - tell your provider what the problems are.
One last note - GTek Technologies Ltd is an Israeli company. You can find our main page here over a very slow connection. This web site is mirrored, at a much faster rate and in the US, here. We have nothing to do with an American company called "GTek Inc." (you can find their site here.
Shachar Shemesh,
sun-sdr@gtek.co.il
I understand it covers the lempel-ziv welsh patent, but is that the only variant covered? What about Lempel-Ziv huffman (LZH)? What exactly is the welsh varient?
The "requirment" was introduced by Bruce and his friends in his twofish proposal. I have not found a trace of this requirment in the original AES requirments.
On page 14 of the TwoFish proposal, titled "Language, Compiler, and Processor Choice", they say "It is clear that the Borland C 5.0 compiler chosen as the standard AES reference is not the best optimizing compiler."...
Optimizing for small hardware was introduced by Bruce and friends, perhaps to distinguish their submission from other submissions?
In order to perform DRoot, you need Euler's Phi, and in order to get that, you need to factor the public key. This is, of course, unless someone finds a better way.
The point is, that someone may find a way to do DRoot, and bypass the factoring, as well as Euler's Phi, problem.
I think we can conclude that factoring complexity >= finding Phi >= finding DRoot.
As for Diffie-Helman, it is based on the difficulty of DLog. The base modulo for DH is a prime number itself. Factoring, therefor, does not enter into it at all.
You tell me how good they are. DES was not publicly reviewed before it was released. How long did it take the rest of the world to find diffrential cryptanalisis (why IBM and the NSA knew about it when DES was invented)?
Now if only Israel came to a similar conclusion, I might actually be able to use those things.