I get 40 mpg in a Toyota Echo, anywhere from $6-9k less than a Prius. Of course on the windswept freeways you get pushed around a bit. So I'm glad its got airbags. Plus I'd hate to crash it. The probability of survival is probably lower than most cars.
I have a question. If I link to libstdc++ from GNU which is GPLv2, in my proprietary C++ code, then does that make my app my default GPL? By the court definition it seems that would be true. If so, that is terrible!
Yahoo Search is actually improving over Google. It is very subtle, but for key information and specific information on some select topics Yahoo is now my second choice if Google loads up a bunch of crap. I remember that this is the way I use to operate with Altavista and Google.
MSN will do well with Dr. Flake. Bring one good guy, he knows how to hire other good guys. That is how Google works, even MSFT. Secondly, the Yahoo website is quite robust.
Web search and Bioinformatics research are hot topics for companies today.
Where do this technologists come from? They have no idea what they are talking about and just spout old school ideas. These ideas have been talked about since the dawn of time. The reason mainframes went away as did terminal processing is that users wanted to freedom and flexibility of their information without having to rely on anyone else.
It will not be a desktop computer. Nobody in their right minds would host all of a users files at the moment. It will be more a web services application framework available for any OS and handheld device. Tivo like devices and VoIP like devices as well.
What does google do well? Store and search information. You already have Google Desktop. Google PC is just stupid. Picasa was probably bought for some other reason than the application being cool. Probably a similar strategy to.Mac and iPhoto, but for PC users. So if you want to publish your information to share with others you use a good service (ad driven)
Having a NT/XP OS designer allows you to WRITE better WINDOWS consumer software. NOT DESIGN an OS. Even if he doesn't write the windows software, he can train and educate others to the hidden APIs. He may want to work on more interesting problems involved in massively large computational systems.
Bioinformatics or Systems Biology. You can go a long way in multiple industries with this training. Comprehensive in engineering, biology, mathematical modeling of dynamics systems, information theory... woot!
Copy protection protects the implementation from copying. I.e. it protects outright theft of the code. The patent protects the process or the algorithm or the ingenious invention. Do we have a right to patent an idea or process when its implementation is in software? Is software any different than a physical (manufacturing or biotech) invention? Is information "patentable". Do you think Google's pagerank deserves patent protection? All it is is a Markov chain with some fancy extra bits. Why is a plugin (a web page with fancy bits) not patentable?
If you invent something, a publically disclose it through the USPTO you have a right to market protection in the USA.
Yes, even 1-click, or a laser toy to entertain cats. Or whatever is reducible to a commericial or competitory product. The question is, what a patents effect will be on open source. Usually academics are not sued for patent infringement when they perform research on a patented idea. You might have to buy a license intrinsic with the product.
Open source per se may not be sueable in a distributed fashion, but certainly people making money of open source inventions are.
Computer software evolves regardless of the patent space. If a patent blocks you, invent a way around it.
The Netflix patent does not affect you or me, it affects Netflix and any big-time competitors. That is the way the patent system was designed. To protect the little guy from big corporate stealing.
For instance you can patent page-rank. It gives you an output of useful information. But you can't patent what you do with that information. But pagerank is a statistic, "a score", or a way of generating scores? You can surely compute your own score using a slightly different model and not be infringing.
A patent represent IP that was an idea put into practice. You disclose the invention publically with hope to profit from your invention. Profiting can mean preventing others from using it, but more generally, what if wanted to license it to profit and MSFT said "screw you" this is obvious, no way.
Anyway, someone who has 20 years experience in the field should be able to implement the idea from the patent. That doesn't mean the idea was obvious at all. 1993, linked objects in a browser? OLE is not prior art? Obvious, nope. Personally, I think the Google patent on page rank was obvious, but that was after reading the patent. Just because you think something is obvious does not mean it is. It is a legal claim determined by patent masters. Patent masters ARE engineerings and scientists.
Patents don't have to be implemented necessarily. And you say that going after a big corp with a patent is a bad thing. Well big corps patent more than anyone else. The patent system is designed to work EXACTLY as it SHOULD do, in the case of ELOAS vs. MSFT.
Just because WE DON'T like PATENTS means ABSOLUTELY nothing.
In my post, "ideas" was quoted... by that I meant that it was a patentable idea, or even "put into practice" -- not just an idea. Putting Viola into practice (if he did), or even storing it on a UC server would in general subsume his prior rights and does not necessarily imply public disclosure. But you are right, If it was in practice at all then public disclosure would invalidate any patentable parts in any other country than the US immediately. In the US you have 1 year to file for a patent after public disclosure. Since the timeline of Pei Wei's alledged disclosure is in agreement with the 1 year filing by the UC over a provisional patent things get murky. And if Pei Wei was a student of UC and his academic agreement said "all your base belong to us".. i.e. any IP he comes up with is the UCs then his prior art is part of the UC invention.
The 1st and 4th point are the same because of emphasis. It actually probably doesn't matter because the rumor is that Pei Wei's "practical" piece didn't work in court. If it's broke, it ain't prior art. That is why his testimony was excluded.
Most Universities have IP claims on your inventions, regardless of whether you are faculty, staff, or student. Same holds in England BTW. I know, I went to a school there and a school here. Same deal. I checked.
BTW Prelimary Applications are SECRET between the USPTO and the INVENTOR... You keep them secret to PREVENT other people from trying to invalidate the patent before it is APPROVED. This is STANDARD PRACTICE and protects your legal rights. Usually they last one year only. MSFT wouldn't know about it.
MSFT lawyers are probably told to spare no expense to get the claim thrown out. If Pei Wei is all they have, they are screwed. HENCE the FUD campaign.
You REALLY need to get your facts straight before you respond.
First, he was a UC student... so if he was ever paid by the UC, in anyway then any "ideas" are University property... regardless of when you invented it (i.e spare time). I am not sure what the academic agreement was... but most Universities (even when you pay to go to school) take your patentable ideas. This is even for undergraduate students. But that might be a new law, rather than an old one.
Second, people usually file a preliminary patent 1 year before the real on gets filed.. In this case it may predate Pei Wei's claims that his prior art existed... (i.e. the main Eloas patent was filed Nov 1994... So 1 year previous was Nov 1993. Which is more or less around the same time as Pei Wei claims to have "shown" his code around. "Showing" your code around generally means nothing. Usually the UC requires an extensive documentation process as well as the "date" the invention was made practical.. That all gets examined by the patent people.
Third, we as the public do not know the specific Court arguments are, so we can't really say what is what.. I am sure the judge had a good reason to restrict the presentation of Viola as prior art.
Fourth, Pei Wei was a UC student... I am pretty sure that discounts the main argument of his art being prior art.
Fifth, big deal if the web breaks. It will evolve quickly. It won't break overnight, people still have to upgrade their browsers. Old apps will still work for awhile. MSFT is just spreading FUD, so is the W3C.
technically its a SFSA which is a type of computer. So no, the journal is not using the term incorrectly. FSA (i.e. regular expressions) are also "computers", but we think of them as programs nowadays. But they technically "compute" a yes/no answer to a decision probably (albeit stochastically).
The Google Cache shows that Aaron developed
a P2P indexing service. Is research now
a crime? Is maintaining an index of publicly
available services a crime? If so, then
google is guilty of theft.
Google Cache of FlatLyn resaerch paper
Record companies, movie companies and any content company pays megabucks for palladium keys to the verisign/msft monopoly so that end user joe schmoe has to buy it. Record companies will pay for this privilege immensely and pass the cost upon consumers... Business offices, no smoking gun yet but signed email, imagine having a prom with your PGP key in it that authenticates against a smart card or a biometric device. No need to remeber the password just a DRM enabled USB key device or even a wireless one. The future is bright for DRM. It won't interfere as DRM will be an add on to all new sound cards and video processors.
The hardware is the solution to stop end to end copying as demonstrated by the successful hacks of most software signatures, and if you use reprogrammable chips you can "update" the signatures. Sure you can mod chip things but you won't be able to play the encrypted content. You might however be able to reverse engineer the encoding scheme from plaintext/cipher text examples.. ie given a song encoded in mp3 then encypt it. Do differential analysis between the two sources, find encryption function + noise. Of course that would be in SEVERE violation of the DMCA and you'd bet you'd get your butt sued ASAP.
Now of course Microsoft will include some noise generator so that it is harder to find the true encyption scheme but given the fact that everything resides in digital format this should be solvable since in the limit of large numbers the noise generator is gaussian. Given multiple samplings you could further refine this. Of course they might sample from a large set of encyrption keys and then of course you have to estimate the total number of keys used from your sampling space for which you would decode them. Good thing the random number generator on NT is broken!!!:) The real point of DRM on the chip is that you cannot conduct this analysis because the end signal is *ANALOG*, you'd have to rencode the signal to digital, and therefore you can never get back to the original digital fidelity. But decoding the ciphertext may still be feasible since the analog rencoding, may be a "good enough" approximation of the digital plaintext. But it may not be if Microsoft includes CONTROL bits in the INAUDIBLE spectrum or NON VISUAL SPECTRUM of the digital signal. Thus the ciphertext analysis would be much more difficult.
Perhaps some Bayesian discrete wavelet estimation, but even that is not enough to detect the control bits. BLAST... DRM is here to stay folks.
Cease continued development. Bill with interest, after 1-2 months report them as an unpaid debtor negatively effecting their business credit. Did you have that in your contract terms?
Also put in a clause that says any non-payment of any owed debt results in the software rights not being transferred from you to them. That is you still own the software. Then sue if they use it (piracy).
If it is a blame game, cut your losses, cease development, then bill them for the work done so far. If they really need it, they'll pay. Also, don't give them any specs. Specs should be bought in advance from you (cash payable on delivery) with 2 price options: 1) if you are developer 2) if they want to use another developer. Contract should stipulate severe penalties for breach of said contract.
Computer consultants seem to loose track of this issue, they forget that the business is a customer and subject to the same rules of credit damage that normal people are. Don't let them get away with it. If you do, it shows your lack of professionalism.
I generally start clients out on a payment plan that is money gets contributed no matter what the progress is, if the company is a good company then they will have no problem with it.
If you do tell a VC do it by word of mouth. Non-traceable. I don't think that is the best idea. Just pursue your legal channels. You might find they actually start to pay when the VC roles in.
Slashdot Mirror
I get 40 mpg in a Toyota Echo, anywhere from $6-9k less than a Prius. Of course on the windswept freeways you get pushed around a bit. So I'm glad its got airbags. Plus I'd hate to crash it.
The probability of survival is probably lower than most cars.
I have a question. If I link to libstdc++ from GNU which is GPLv2, in my proprietary C++ code, then does that make my app my default GPL? By the court definition it seems that would be true.
If so, that is terrible!
Yahoo Search is actually improving over Google. It is very subtle, but for key information and specific information on some select topics Yahoo is now my second choice if Google loads up a bunch of crap. I remember that this is the way I use to operate with Altavista and Google.
MSN will do well with Dr. Flake. Bring one good guy, he knows how to hire other good guys. That is how Google works, even MSFT. Secondly, the Yahoo website is quite robust.
Web search and Bioinformatics research are hot topics for companies today.
Where do this technologists come from? They have no idea what
they are talking about and just spout old school ideas. These ideas have been talked about since the dawn of time. The reason mainframes went away as did terminal processing is that users wanted to freedom and flexibility of their information without having to rely on anyone else.
It will not be a desktop computer. Nobody in their right minds
would host all of a users files at the moment. It will be more
a web services application framework available for any OS
and handheld device. Tivo like devices and VoIP like devices as well.
What does google do well? Store and search information. You already have Google Desktop. Google PC is just stupid. Picasa was probably bought for some other reason than the application being cool. Probably a similar strategy to
Having a NT/XP OS designer allows you to WRITE better WINDOWS consumer software. NOT DESIGN an OS. Even if
he doesn't write the windows software, he can train and educate
others to the hidden APIs. He may want to work on more interesting problems involved in massively large computational systems.
Oh baby, major lawsuit on this one. Harvard is making a big
mistake. The 119 person class action lawsuit will be easily won.
Bioinformatics or Systems Biology. You can go a long way in multiple industries with this training. Comprehensive in engineering, biology, mathematical modeling of dynamics systems, information theory... woot!
Ummm... Halo 3, duh!
Copy protection protects the implementation from copying. I.e. it protects outright theft of the code. The patent protects the process or the algorithm or the ingenious invention. Do we have a right to patent an idea or process when its implementation is in software? Is software any different than a physical (manufacturing or biotech) invention? Is information "patentable". Do you think Google's pagerank deserves patent protection? All it is is a Markov chain with some fancy extra bits. Why is a plugin (a web page with fancy bits) not patentable?
If you invent something, a publically disclose it through the USPTO you have a right to market protection in the USA.
Yes, even 1-click, or a laser toy to entertain cats. Or whatever is reducible to a commericial or competitory product. The question is, what a patents effect will be on open source. Usually academics are not sued for patent infringement when they perform research on a patented idea. You might have to buy a license intrinsic with the product.
Open source per se may not be sueable in a distributed fashion, but certainly people making money of open source inventions are.
Computer software evolves regardless of the patent space. If a patent blocks you, invent a way around it.
The Netflix patent does not affect you or me, it affects Netflix and any big-time competitors. That is the way the patent system was designed. To protect the little guy from big corporate stealing.
For instance you can patent page-rank. It gives you an output of useful information. But you can't patent what you do with that information. But pagerank is a statistic, "a score", or a way of generating scores? You can surely compute your own score using a slightly different model and not be infringing.
A patent represent IP that was an idea put into practice. You disclose the invention publically with hope to profit from your invention. Profiting can mean preventing others from using it, but more generally, what if wanted to license it to profit and MSFT said "screw you" this is obvious, no way.
Anyway, someone who has 20 years experience in the field should be able to implement the idea from the patent. That doesn't mean the idea was obvious at all. 1993, linked objects in a browser? OLE is not prior art? Obvious, nope. Personally, I think the Google patent on page rank was obvious, but that was after reading the patent. Just because you think something is obvious does not mean it is. It is a legal claim determined by patent masters. Patent masters ARE engineerings and scientists.
Patents don't have to be implemented necessarily. And you say that going after a big corp with a patent is a bad thing. Well big corps patent more than anyone else. The patent system is designed to work EXACTLY as it SHOULD do, in the case of ELOAS vs. MSFT.
Just because WE DON'T like PATENTS means ABSOLUTELY
nothing.
Dear Anonymous Coward,
In my post, "ideas" was quoted... by that I meant that it was a patentable idea, or even "put into practice" -- not just an idea.
Putting Viola into practice (if he did), or even storing it on a UC server would in general subsume his prior rights and does not
necessarily imply public disclosure. But you are right, If it was in practice at all then public disclosure would invalidate any patentable parts in any other country than the US immediately. In the US you have 1 year to file for a patent after public disclosure. Since the timeline of Pei Wei's alledged disclosure is in agreement with the 1 year filing by the UC over a provisional patent things get murky. And if Pei Wei was a student of UC and his academic agreement said "all your base belong to us".. i.e. any IP he comes up with is the UCs then his prior art is part of the UC invention.
The 1st and 4th point are the same because of emphasis. It actually probably doesn't matter because the rumor is that Pei Wei's "practical" piece didn't work in court. If it's broke, it ain't
prior art. That is why his testimony was excluded.
Most Universities have IP claims on your inventions, regardless of whether you are faculty, staff, or student. Same holds in England BTW. I know, I went to a school there and a school here. Same deal. I checked.
BTW Prelimary Applications are SECRET between the USPTO and the INVENTOR... You keep them secret to PREVENT other people from trying to invalidate the patent before it is APPROVED. This
is STANDARD PRACTICE and protects your legal rights. Usually they last one year only. MSFT wouldn't know about it.
MSFT lawyers are probably told to spare no expense to get the claim thrown out. If Pei Wei is all they have, they are screwed.
HENCE the FUD campaign.
You REALLY need to get your facts straight before you respond.
First, he was a UC student... so if he was ever paid by the UC, in anyway then any "ideas" are University property... regardless of when you invented it (i.e spare time). I am not sure what the academic agreement was... but most Universities (even when you pay to go to school) take your patentable ideas. This is even for undergraduate students. But that might be a new law, rather than an old one.
Second, people usually file a preliminary patent 1 year before the real on gets filed.. In this case it may predate Pei Wei's claims that his prior art existed... (i.e. the main Eloas patent was filed Nov 1994... So 1 year previous was Nov 1993. Which is
more or less around the same time as Pei Wei claims to have "shown" his code around. "Showing" your code around generally means nothing. Usually the UC requires an extensive documentation process as well as the "date" the invention was made practical.. That all gets examined by the patent people.
Third, we as the public do not know the specific Court arguments are, so we can't really say what is what.. I am sure the judge had a good reason to restrict the presentation of Viola as prior art.
Fourth, Pei Wei was a UC student... I am pretty sure that discounts the main argument of his art being prior art.
Fifth, big deal if the web breaks. It will evolve quickly. It won't break overnight, people still have to upgrade their browsers. Old apps will still work for awhile. MSFT is just spreading FUD, so is the W3C.
Sixth, I have no stake in this either way.
technically its a SFSA which is a type of computer. So no, the journal is not using the term incorrectly. FSA (i.e. regular expressions) are also "computers", but we think of them as
programs nowadays. But they technically "compute" a yes/no answer to a decision probably (albeit stochastically).
The Google Cache shows that Aaron developed a P2P indexing service. Is research now a crime? Is maintaining an index of publicly available services a crime? If so, then google is guilty of theft. Google Cache of FlatLyn resaerch paper
Record companies, movie companies and any content company pays megabucks for palladium keys to the verisign/msft monopoly so that end user joe schmoe has to buy it. Record companies will
:) The real point of DRM on the chip is that you cannot conduct this analysis because the end signal is *ANALOG*, you'd have to rencode the signal to digital, and therefore you can never get back to the original digital fidelity.
pay for this privilege immensely and pass the cost upon consumers... Business offices, no smoking gun yet but signed email, imagine having a prom with your PGP key in it that authenticates against a smart card or a biometric device. No need to remeber the password just a DRM enabled USB key device or even a wireless one. The future is bright for DRM. It won't interfere as DRM will be an add on to all new sound cards and video processors.
The hardware is the solution to stop end to end copying as demonstrated by the successful hacks of most software signatures, and if you use reprogrammable chips you can "update" the signatures. Sure you can mod chip things but you won't be able to play the encrypted content. You might however be able to reverse engineer the encoding scheme from plaintext/cipher text examples.. ie given a song encoded in mp3 then encypt it. Do differential analysis between the two sources, find encryption function + noise.
Of course that would be in SEVERE violation of the
DMCA and you'd bet you'd get your butt sued ASAP.
Now of course Microsoft will include some noise generator so that it is harder to find the true encyption scheme but given the fact that everything resides in digital format this should be solvable since in the limit of large numbers the noise generator is gaussian. Given multiple samplings you could further refine this. Of course they might sample from a large set of encyrption keys and then of course you have to estimate the total number of keys used from your sampling space for which you would decode them. Good thing the random number generator on NT is broken!!!
But decoding the ciphertext may still be feasible since the analog rencoding, may be a "good enough" approximation of the digital plaintext. But it may not be if Microsoft includes CONTROL bits in the INAUDIBLE spectrum or NON VISUAL SPECTRUM of the digital signal. Thus the ciphertext analysis would be much more difficult.
Perhaps some Bayesian discrete wavelet estimation, but even that is not enough to
detect the control bits. BLAST... DRM is here to stay folks.
What do you think of openmosix? Is it the "true"
answer for real beowulf computing?
What progress is being made in architecting security in the Internet in regards to protocol 'attacks' and global traffic shaping?
Any comments on the Whitehouse Cybersecurity proposal released last week?
they forget that a lot of the community
exists b/c of microsoft: Microsoft licensing, unfair tactics, and lack of innovation and transparency.
Sure the build a better desktop but they
build a crappy server. And even more so,
why do thing their way when you can do it
your way?
will it survive the ./ effect??
Cease continued development. Bill with interest, after 1-2 months report them as an unpaid debtor negatively effecting their business credit. Did you have that in your contract terms?
Also put in a clause that says any non-payment of
any owed debt results in the software rights
not being transferred from you to them. That is you still own the software. Then sue if they use it (piracy).
If it is a blame game, cut your losses, cease development, then bill them for the work done
so far. If they really need it, they'll pay.
Also, don't give them any specs. Specs should
be bought in advance from you (cash payable on delivery) with 2 price options: 1) if you are developer 2) if they want to use another developer. Contract should stipulate severe penalties for breach of said contract.
Computer consultants seem to loose track of this issue, they forget that the business is a customer and subject to the same rules of credit damage that normal people are. Don't let them
get away with it. If you do, it shows your lack
of professionalism.
I generally start clients out on a payment plan that is money gets contributed no matter what the progress is, if the company is a good company then
they will have no problem with it.
If you do tell a VC do it by word of mouth. Non-traceable. I don't think that is the best idea.
Just pursue your legal channels. You might find they actually start to pay when the VC roles in.
as annoying as the old html command:
blink