5 ms·

If you read the claims they are apparently infringing upon, it's all math with a thin veneer of "on a magnetic storage medium". It's literally mathematical sign

by pi18n 14y ago

If you read the claims they are apparently infringing upon, it's all math with a thin veneer of "on a magnetic storage medium". It's literally mathematical signal processing. I would bet a few dozen bucks that it is obvious to anyone studying signal processing, since it seems like an advancement on a Kalman filter.

monochromatic 14y ago
Yeah, we're probably doing a better analysis glancing at the claims than the team of patent attorneys defending against a billon dollar lawsuit did.
wmf 14y ago
Yeah, it's so obvious that it was published in IEEE Transactions on Information Theory and the author got a PhD from CMU for it.
- Brashman 14y ago
  I agree. Since this is coming out of research out of a university, I would be more inclined to guess that (at the time at least) this was actually innovative and new work. The fastest way for a paper to be rejected is for it to have been done already, or even for it to be "obvious to anyone" with basic knowledge in the field. Also, I don't see why the fact that it's advancing existing techniques or that it's just math somehow makes it less important. There are whole fields that are just math, such as computer science originally.
  - pi18n 14y ago
    I don't want to make math and computer science sound unimportant -- it's an important distinction because pure math is not supposed to be patentable. But for whatever reason if you include a hypothetical machine it is suddenly patentable.
- pi18n 14y ago
  So I lose the bet, but it's still mathematics that shouldn't be patentable in the first place.
uvdiv 14y ago
Not a Kalman filter, but rather a Viterbi decoder -- a type of error-correcting code. https://en.wikipedia.org/wiki/Viterbi_decoder https://en.wikipedia.org/wiki/Viterbi_decoder From lurking on HN, I read that the whole field of ECC's is infested with patents. There's even an example on here of a specific startup (allegedly) destroyed by an ECC patent: http://news.ycombinator.com/item?id=3432322 http://news.ycombinator.com/item?id=3432322

by pi18n 14y ago

monochromatic 14y ago

Yeah, we're probably doing a better analysis glancing at the claims than the team of patent attorneys defending against a billon dollar lawsuit did.

wmf 14y ago

Yeah, it's so obvious that it was published in IEEE Transactions on Information Theory and the author got a PhD from CMU for it.

Brashman 14y ago
I agree. Since this is coming out of research out of a university, I would be more inclined to guess that (at the time at least) this was actually innovative and new work. The fastest way for a paper to be rejected is for it to have been done already, or even for it to be "obvious to anyone" with basic knowledge in the field. Also, I don't see why the fact that it's advancing existing techniques or that it's just math somehow makes it less important. There are whole fields that are just math, such as computer science originally.
- pi18n 14y ago
  I don't want to make math and computer science sound unimportant -- it's an important distinction because pure math is not supposed to be patentable. But for whatever reason if you include a hypothetical machine it is suddenly patentable.
pi18n 14y ago
So I lose the bet, but it's still mathematics that shouldn't be patentable in the first place.

uvdiv 14y ago

Not a Kalman filter, but rather a Viterbi decoder -- a type of error-correcting code. https://en.wikipedia.org/wiki/Viterbi_decoder https://en.wikipedia.org/wiki/Viterbi_decoder From lurking on HN, I read that the whole field of ECC's is infested with patents. There's even an example on here of a specific startup (allegedly) destroyed by an ECC patent: http://news.ycombinator.com/item?id=3432322 http://news.ycombinator.com/item?id=3432322