Could still be dangerous if the steganography is detectable
If a regime decides to continue to allow Flickr despite this, it sounds like they could use Collage themselves to detect and decode the hidden material. Once they know which pictures contain censored information they log any downloads of those images against the user's IP address and use that as information about who in their own population is reading it. I would want the program to need the right key to even be able to detect that there is hidden material present before I used something like this.
There has to be a connection with atmospheric pressure, the professor is not completely right. A siphon cannot work if the hump is significantly larger than the equivalent height of the atmospheric pressure for the liquid, i.e. about 34 feet for water at 1 atm, 30 inches for mercury. It is the pressure of the atmosphere that is pushing the water up the pipe, although gravity does pull it down the lower side causing the liquid to flow. If you try to make the hump any bigger than that height, you'll get a void (vacuum) forming at the top of the hump and the flow will stop.
The Wikipedia entry for Atmospheric Pressure says "This is also the maximum height to which a column of water can be drawn up by suction" which is exactly what a siphon does, it uses gravity to generate suction that pulls the liquid over the hump. Surface tension might have some effect on the maximum height of the hump, but it couldn't be very big and would depend on the diameter of the pipe.
Do any of the mobile browsers support a View Source option? I'm guessing not, as that would only be needed by people wanting to use the devices for creative purposes; the functionality is irrelevant for consumptive users. This supports Tim Wu's concerns to a small extent.
The FTC really didn't think this through when they set up the site in the first place; if they has used a .gov address instead it would have made it much easier for people to recognize when they're using the right address.
If it was Homeland Security who claimed that the screener would be at a remote location, that promise can only apply to machines installed at US airports; the TSA have no jurisdiction over Heathrow Airport. The British have their own rules, which include that they can't use the scanners on children since the resulting images would then be classified as child pornography.
I'm no lawyer and wouldn't know how to start searching for such things, but I suspect the legal status of works made by owned animals might be similar to that of works created by owned people, meaning by slaves. Of course there may never have been any such cases, but if there were they might give some guidance...
If remastering means to going back to the original studio multi-track recordings (digitizing as necessary) and re-mixing them, I could see that as being a different kind of thing (and somewhat more worthy of a new copyright for a record company which actually does that) than just re-encoding the digital audio data which is what it sounds like Bluebeat were trying to do. As a result I'm highly skeptical that a court would see these two issues as the same (but IANAL so what do I know).
I sing in a professional choir; while we do perform music composed by our director, in most cases the stuff we sing is by living composers, whose publishers all charge per-copy royalties to record their songs. We couldn't afford to pay those fees ourselves on copies that we give away for free, so those recordings are not infinite goods. I imagine that any band recording cover songs that are not yet out of copyright would have the same problem, but I've never heard this being discussed.
I would love to be able to put up a whole bunch of our past performances as MP3s on our website, but I can only do that legally for the limited amount of our output where the compositions are no longer in copyright. Any ideas?
No, they never created the logs in the first place.
IIRC the ARS Technica coverage of this story made it clear (they actually talked to Bahnhof) that the ISP never actually logged the information in the first place, and they're not deleting logfiles now. I don't think they've changed what they do at all, they're just bringing this fact to the attention of potential customers who might be interested in their lack of logging.
The idea that copyrights should die with the author is obviously a bad one, because as others have pointed out it creates an incentive for others to take advantage of that by killing the author. I think most people would agree that such incentives would be a bad thing, and it's actually a disincentive to the author to create new works because with each success their death becomes worth more to those people who would want to exploit it.
A part of an author's incentives to produce new works may also be to provide for their heirs. An established author who is older or dying would have little or no incentive to create new works if they would be unable to pass on the value of those works to their heirs.
Therefor the idea that copyrights should not be inheritable is obviously wrong from the perspective of the incentives of authors.
However, extending copyright terms is completely pointless as far as incentives go if the new terms apply retroactively to works that already exist at the time the law is enacted. Doing so would merely be giving the value of the future public domain lost over to the current owners for no public gain. Any newly enacted terms would apply to any new works and thus provide the increased incentives to create, but the retroactive provision is a pure negative for the public.
I agree with many that the existing copyright terms are too long, but that's not really the topic here.