RabidMongoose
Lifer
- Aug 14, 2001
- 11,061
- 0
- 0
Originally posted by: Roger
No one can copy your invention/process for seven years without paying you royalties.
I think that it's usually 17 years.
Originally posted by: Roger
No one can copy your invention/process for seven years without paying you royalties.
Originally posted by: PhasmatisNox
My dad got a patent. It cost about $6000 for a patent lawyer, although I have the feeling that he got pwned.
Originally posted by: badmouse
The four-year wait for a patent is a great thing, because "patent pending" is equally good protection - in fact, it's better protection than an actual patent because you don't have to disclose anything. A patent, once approved, is released to the public.
Originally posted by: sonambulo
Originally posted by: badmouse
The four-year wait for a patent is a great thing, because "patent pending" is equally good protection - in fact, it's better protection than an actual patent because you don't have to disclose anything. A patent, once approved, is released to the public.
Is there any way to look at a patent app while it's pending? basically my situation is this: homeboy's got a patent pending stamped on all of his products. they're pretty much a monopoly in a niche (read:embarrassing) market. i need to see what he's got patented so i can develop a workaround but he's being tightlipped. what can i do?
Originally posted by: viking1966
one way to protect your idea while you go through the process is to mail yourself the plans and never open it.. this way you have a government post/stamped and data record of your idea that will hold up in court if someone tries to steal your idea..
Originally posted by: sonambulo
from what i understand there's a test or tests to determine of a product is a new and unique piece. anyone know where i can find this test?
Originally posted by: SagaLore
Originally posted by: theNEOone
hmm, that's cool. my school does spam blocking/filtering based on hundreds of rules, assigns a weight to each of these rules, and then finally determines that likelihood that a particular e-mail is spam. if it's more than 50% likely that the e-mail is spam, it adds a tag to the beginning for the sujbect [PMX:###] to indicate that it's spam, with each "#" indicating an extra 10% probability that the e-mail is spam. (therefore, the above example indicates an 80% chance that the email is spam). then through outlook you can add a simple rule to filter out these e-mails based on percentage. i have really relaxed rules - i'm filtering at 50% probability. the rules are damn good, so far i get 100% spam blocking, and maybe 1-2% false positives. i can probably increase the outlook rules to filter out only 60-70, and i might get rid of the false positives, but it works so well that i think that i'll just leave it as it is.Originally posted by: SagaLore
Originally posted by: theNEOone
what does your process accomplish?
=|
It stops a degree of spam, accurately.
here's some possible competition. my school uses PureMessage
=|
My antispam method is only complimentary, I don't think it would work well by itself.
What your school is using, PureMessage, is a bayesian algorithm designed by ActiveState which was recently acquired by Sophos.
The difficulty with bayesian is that there are hundreds, sometimes thousands of tokenized rules. The value of each token is adjusted based on email traffic. This is known as "bayesian training". Unfortunately, where this succeeds as a desktop solution, fails as an enterprise solution. When you have 500 employees, each within departments of different themes, the bayesian loses effectiveness while increasing false positives.
Another problem with Bayesian is that spammers quickly identify the rulesets, and use their own tokenizing software to counteract it. It's war of the algorithms. That's why lately you see emails that use non traditional characters in their content, or a dump of an excerpt of some book, or no text at all and just load an image via html.