C'mon, that's so Web 1.0! What about searching for pr0n by drawing a sketch? Now that would be cool! (On a more serious note, there's on-going research on that topic.)
Application of the two-stage photosynthesis and
H2 production protocol to a green alga mass
culture could provide a commercially viable method of renewable
hydrogen generation. Table I provides preliminary estimates of maximum possible yield of
H2 by green algae, based on the luminosity of the
sun and the green algal photosynthesis characteristics. Calculations
were based on the integrated luminosity of the sun during a cloudless
spring day. In mid-latitudes at springtime, this would entail delivery
of approximately 50mol photons m-2
d-1 (Table I, row 1). It is generally accepted
that electron transport by the two photosystems and via the hydrogenase
pathway for the production of 1mol H2 requires
the absorption and utilization of a minimum of 5mol photons in the
photosynthetic apparatus (Table I, row 2). On the basis of these
"optimal" assumptions, it can be calculated that green algae could
produce a maximum of 10mol (20g) H2 perm2culture area per day. If yields of such
magnitude could be approached in mass culture, this would constitute a
viable and profitable method of renewable H2
production.
However, this optimistic scenario cannot be realized with present day
know-how. Three biologically "gray areas" directly impact this
H2 production technology. (a) The yield of
H2 production currently achieved in the
laboratory corresponds to only 15% to 20% of the measured capacity of
the photosynthetic apparatus for electron transport (Melis et al.,
2000). (b) The optical properties of light absorption by green algae
impose a limitation in terms of solar conversion efficiency in the alga
chloroplast. This is because wild-type green algae are equipped with a
large light-harvesting chlorophyll antenna size to absorb as much
sunlight as they can. Under direct and bright sunlight, they could
waste up to 60% of the absorbed irradiance (Neidhardt et al., 1998;
Melis et al., 1999). This evolutionary trait may be good for survival
of the organism in the wild, where light is often limiting, but it is not good for the photosynthetic productivity of a green algal mass
culture. This optical property of the cells could further lower the
productivity of a commercial H2 production farm.
(c) The current necessity to cycle a culture between the two stages (normal photosynthesis in the presence of S alternating with
H2 production upon S deprivation) introduces a
"down time" as far as H2 production is
concerned. It is inevitable that the "down time" would further
erode the yield of the H2 production process. Thus, with current technology, it is estimated that the actual yield of
H2 production would be lower than that of the
theoretical maximum shown in Table I, achieving perhaps a mere 10%, or
lower, than the calculated theoretical maximum. It is clear that these three specific biological challenges (a-c) need to be overcome to
effect greater actual yields of green alga H2 production.
Well, I'm quite sure they wouldn't like us/.ers. Unshaven, not showered for weeks... Looks like we sure do have an evolutionary advantage with respect to the not/. population.
But then, they wouldn't find us anyway, hidden away in the basement, armored with two racks and and way too large hoody.
As I've said elsewhere, I guess this scheme is mainly targeted at purchases with content delivered via the network. Not CDs you buy at five-and-dime's.
So what? You people have to get your mind together. First you complain about the bundling of WebKit with OS X. Then we point out that it is open source and can be replaced at will. Now you complain that Mac OS X is not open source. How does that support your initial claim about the bundling?? You can remove WebKit whenever you like. Heck, the binaries are even conveniently packaged together in a directory, so remove it, and plug in your own build of WebKit, or even something completely new but with a compatible API.
WebKit is the system framework used on Mac OS X by Safari, Dashboard, Mail.app, and many other OS X applications. It is based on the KHTML engine from KDE.
There are many ways in which you can get involved with the project. You can either check out and build the source code or download the latest nightly build to try out the latest version of WebKit in your Safari browser. Once you have either of these, you can report bugs you find in the software. You can help out with bugs by providing reductions or by submitting patches for review. Contributors with a proven track record of good patches will be given check-in access to the repository.
BTW, which license is KHTML and how is it they get away not opening MacOS X? I'd like to know the details.
Well, but if you get 5000 different watermarked copies, this also means that there will be 5000 users sued for distributing them in the first place.
And even if you know the algorithm, what does this help you with a different song? Remember, each watermark is different, based on a random string. Compare it to an encryption algorithm. You can think of the watermark as a section of the original file M, encrypted with some key K (which, BTW, is a completely random string only known by the entity which produced the file). What you now basically have is a known ciphertext C and the algorithm. Now what? From this, you can neither deduce the key K nor the plaintext M. Having several ciphertexts C1, C2, C3 etc. at your disposal may help, but keep in mind that they are all encrypted with different keys K1, K2, K3, etc., as opposed to a normal attack where you have several ciphertexts C1, C2, C3 etc., but all encrypted with the same key K.
Either way, I don't see them putting unique ID codes on mass-market CDs anytime soon. Imagine the size of the tracking database - and the distribution chain cooperation required. Instead they'll make one version for each distribution channel or region, and use that as part of the evidence in building a case ("we know the CD was sold in Nebraska - which is exactly where our suspect lives!").
Well, it seems to be mainly targeted at files purchased online and delivered via the network. In this case it is trivial to watermark each distributed file individually.
Doesn't matter how smart they are. The 1ee7 trading group gets 10 copies of the CD, averages the values at each sampling interval, and the watermark is gone (or obscured beyond use).
Above said, assume for now we are talking about Internet-delivered content. If you are a direct member of this trade group, and you can be sure that the copy you yourself purchased is going to be scrambled like this and does not appear anywhere else, then you don't have to be worried. But as a normal p2p filesharer, as soon as some RIAA/MPAA guy gets a hand at your file anywhere else than your computer, you're hosed immediately. I guess this will definitely scare some people off of filesharing.
Preach it brother, preach it!! Those were the golden days! I remember to once have written a parallelised version of the startup-sequence, running all dependency-free commands in the background. Though I forgot if it gained me a lot of performance.
And you, sir, have never heard of Steganography, or Hamming codes. Do you really think they needed scientists to come up with a scheme like you proposed? Haha, these people are, let me guess, 100x smarter than you. Sorry to have to say it.
Yes, but the intent is to scare people to start spreading a file at all. If they only detect one single illegal copy, they don't even have to go through the hassle of looking at network logs to trace back the route how it came to them, they can simply come to your door and sue you for illegaly sharing files.
If, as of today, you only share it with few people, or take care of the secrecy of the paths it is distributed on, you can basically sleep calm at night. But if you spread a watermarked file, they can directly identify you as the distributor, no matter wheter it went through 100 different computers and users, crossed several borders, and went one time around the earth, until it ended up at the lawyer.
I guess yes, as every other thing does. Think about credit card fraud or digital identity theft. It all does not have "physical" evidence, but only digital. Of course you could always make a printout...
But you start to have problems if the watermarks in each file overlap. Then you need more than two files.
Also think about the situation where files all have their watermarks at the same position. If the watermark is an error correcting code, and the Hamming distance between the valid strings varies enough, you might even be able to find out which two copies were used to provide the "cleansed" copy.
Slashdot Mirror
Looks like your name doesn't fit your attitude well!
But then, they wouldn't find us anyway, hidden away in the basement, armored with two racks and and way too large hoody.
Or how about web-based solutions?
LGPL.
And even if you know the algorithm, what does this help you with a different song? Remember, each watermark is different, based on a random string. Compare it to an encryption algorithm. You can think of the watermark as a section of the original file M, encrypted with some key K (which, BTW, is a completely random string only known by the entity which produced the file). What you now basically have is a known ciphertext C and the algorithm. Now what? From this, you can neither deduce the key K nor the plaintext M. Having several ciphertexts C1, C2, C3 etc. at your disposal may help, but keep in mind that they are all encrypted with different keys K1, K2, K3, etc., as opposed to a normal attack where you have several ciphertexts C1, C2, C3 etc., but all encrypted with the same key K.
Well, it seems to be mainly targeted at files purchased online and delivered via the network. In this case it is trivial to watermark each distributed file individually.
Doesn't matter how smart they are. The 1ee7 trading group gets 10 copies of the CD, averages the values at each sampling interval, and the watermark is gone (or obscured beyond use).Above said, assume for now we are talking about Internet-delivered content. If you are a direct member of this trade group, and you can be sure that the copy you yourself purchased is going to be scrambled like this and does not appear anywhere else, then you don't have to be worried. But as a normal p2p filesharer, as soon as some RIAA/MPAA guy gets a hand at your file anywhere else than your computer, you're hosed immediately. I guess this will definitely scare some people off of filesharing.
If, as of today, you only share it with few people, or take care of the secrecy of the paths it is distributed on, you can basically sleep calm at night. But if you spread a watermarked file, they can directly identify you as the distributor, no matter wheter it went through 100 different computers and users, crossed several borders, and went one time around the earth, until it ended up at the lawyer.
Also think about the situation where files all have their watermarks at the same position. If the watermark is an error correcting code, and the Hamming distance between the valid strings varies enough, you might even be able to find out which two copies were used to provide the "cleansed" copy.