Slashdot Mirror

If you are looking for more detailed information, along with equasions, here is a link to one of their recent publications on the topic: http://www.cs.dartmouth.edu/~farid/publications/sp 04.html

The article says the research is founded by the Department of Homeland Security. That means that despite the many useful possible aspects of this technology, it will probably never see the light of day. If the DHS is going to be using this technology to identify faked photos, it would be greatly in their interest for the full algorithm and its implementations to not be made available to the public-- since after all if people know what the DHS is using to determine faked photos, they can target the algorithm.

Now that I have written that, looking around, it appears I am actually wrong. If you look at Mr. Farid's personal page, it appears he will be publically presenting a paper covering the fakeness-detection algorithm. I hope the full algorithm will be presented to the academic community.

i checked up on it, and it looks like we're both right:

there's a nice section with diagrams about foldover and aliasing on dartmouth's ea music site.

we are indeed talking about the same thing -- my "foldover" happens if frequencies over the nyquist rate get into the system during the sampling or synthesis process -- aliasing seems to be the same phenomenon when you downsample something, i.e. from 88.2 to 44.1khz.

this stuff is really interesting, i think. :) i'd be interested to see spectral analysis of a signal sampled at 44.1 vs. one that was recorded higher and then downsampled.

as for audio formats -- my home protools system can do 16 or 24 bit I/O up to 48khz to the digital busses. the newer one where i go to school is an HD system and will do I/O up to 48-bit 192khz. my understanding is that if you record onto ADAT or a DA-88 multitracker at 16-bit, you should import it into the digital editor at 24 or higher so the processing and editing you do will retain the fidelity of the original recording.

i'm an analog junkie myself, but i agree that these "new" formats are going to take a while to catch on. i think computer-based distribution is the way to go, even though CDs are pretty convenient. with mp3, aac, ogg, etc. you can pretty much choose the fidelity you want to get out of it and encode it as you see fit. i'd love to hear some classical recordings at a high bit depth and sampling rate just to see if there is a difference... but, for listening to stuff in the car it won't matter.

-matt

I have been doing a high school science research class project on stopping the spreading of internet-borne worms though analysis of epidemic models and such. I have come across many different methods for stopping the distribution of vulnerability-based worms, so I'll share here (in order from most innovative to most obvious): First, a very ingenious method coming from Dartmouth's Institute for Security Technology Studies. They propose a method called monitoring the internet for plumes of ICMP unreachable messages. Software is installed on routers which records the ICMP unreachable messages being sent and sends data every once in a while to a central server which analyzes the data and sees which things are probably random-scanning worms. This is probably the best idea I've seen yet, but most likely the hardest to implement (as router software is usually tried to keep air-tight). The bad ports and such would then be filtered or turned off as appropriate. A second method which may have been talked about on here or not is "good" worms. Worms which sit around and listen for worm data would then send a copy of itself from the computer which was scanning them, therefore fixing another hole and having that computer be another "good" computer. The bad thing with this is that it will only really work when the worm is at its peak, when damage has already been done. It would be useful for cleanup, but of course there are issues with privacy and control would be rampant. Another "solution" is getting users to install firewalls and anti-virus software but thats a more obvious and hard to implement solution. I am modeling all of these possibilities using a mathematical model for epidemics, and seeing where which one would theoretically be most useful and such, and I'll take a look at the method used in the article.

Personally, I agree with you that using your own words more than once is acceptable -- hell, maybe even commendable (laziness is a virtue, right?). Furthermore, it might seem that re-using an old essay (or paragraph thereof) does not meet your dictionary definitoin of "plagiarizing".

However, I know that at my alma mater, we were explicitly told when we matriculated that using your own work a second time was plagiarism under the school's definition (the definition actually said it was okay to use a paper a second time if you got the permission of the professors of both classes in which the paper was used).

So trying to say the school is dumb because they don't conform to your dictionary's definition of a pretty complex legal and ethical issue is a little ingenuous and ignores the fact that this kid may have -- in fact, likely did -- break not only the stated rules but the established standards of collegiate honesty.

Leaving aside the question of whether such a thing is possible

Possible? It has been done.

I think the poster is confusing using quantum codes (first demostrated in 1991, currently commercially available) with breaking codes with quantum computers (still hugely theoretical).

My Red Herring link is now 404 - I had it bookmarked as "The Xbox could lose $900M over eight years" and "Microsoft takes heavy losses on the Xbox". I also have this PDF which does give a cost estimate on the Xbox, but not broken down.

I don't know if the 404 at Red Herring is temporary - here is the link that used to work anyway:

http://www.redherring.com/insider/2002/0624/xbox 06 2402.html

This may relate in someway to how Airline Carriers are rated. Faster turn around time means your aircraft is in flight more, which means it's more efficiently utilized (which makes more $$$). For example, Southwest has the fastest turn around time in the industry (20 minutes): [http://mba.tuck.dartmouth.edu/pdf/2002-2-0012.pdf ]

Here's a nostalgia-filled site dedicated to the Dartmouth Time-Sharing System A bit out of data, but interesting to see the BASIC compiler in some GE assembly language from the early sixties.

Here's a nostalgia-filled site dedicated to the Dartmouth Time-Sharing System A bit out of data, but interesting to see the BASIC compiler in some GE assembly language from the early sixties.

Thomas Kurtz is still alive. Unfortunately John Kemeny is not.

because you wouldn't know which photons contain the data. as soon as you touch it, the other end knows it's datastream has been tampered with.

This is a good overview.

This backup system is fire-proof, though it can be degraded by smoke and fog.

Up here in the Great White North we had a bit of a pinkeye outbreak a few years ago. There was a particularly virulent strain that made its way onto campus, and spread like crazy via all the public computer keyboards. Word on the street was that between half and two-thirds of the campus might have had pinkeye that winter. The CDC even sent some people up to study it. Just goes to show what a few dirty keyboards can do.

poor sco..

*sniffle*

I don't understand what's the big deal here. Hany Farid, at Dartmouth University, has developed a series of successful techniques that help a user tell whether an image is realistic or photorealistic (as in touched up). Hany employs some interesting statistical techniques to tell whether or not a digital signal (be it a photograph, audio recording, etc) has been tampered with or not. Check out his website for more information.

When I was in college at Dartmouth there was a policy that the regularly-made backups of the mail servers were deleted after *one week*. They were daily or weekly backups, can't remember. The given reason for the policy was that it was in the college's interests to have nothing to hand over if there were ever a subpoena.

Here's an anecdote which, for me, sums up all of the differences between Christianity and Islam: I'm sure you're aware of the story from the haddith of the adulterous woman who was brought before Mohammed. She was pregnant with another man's child, so he ordered her to brought back after the child was born. So it was done. Then, he ordered her to be brought back after the child had been weaned. Again, it was done as he ordered. The third time she appeared before him, he ordered her stoned to death. Sure enough, she was executed for having slept with a man not her husband.

For those of you looking for something a little more recent (2002), Robert Douglas Bruce III has a short paper on minimizing bone loss for you to read on your long trip. Lots of good references at the end.

I wonder if they've talked to this guy

He claims to have a system which can detect modifications to photographic images.

Any tampering with a photographic image causes detectable statistical changes. These changes can indicate that the image may have been edited to change the content or possibly that steganographic data has been added.

The Bell Curve is widely loathed because it's wretched science with a nasty, right-wing agenda. I recommend Stephen Jay Gould's rebuttal, one version of which is his review of the book in the New Yorker.

Eh? Not at all. The "roman thing" is quite well documented.

I should know better than to respond to somebody who quotes Fox News as a source on science (particularly when the opening paragraph contains the words "junk," "science," and "environmental" -cheerleaders for the smokestack lobby), but if you feel so good about DDT, why not try sprinkling some on your morning breakfast cereal?

First, Rachel Carson is not a researcher, so Fox's refuting of her writings, and her interpretation of one of the researchers she quotes, doesn't address the large body of research regarding correlations between DDT and its effects on wildlife.

On the subject securing optical links, quantum crypto is an interesting aproach. It is not useful to transmit a lot of data, but can be used in secure key interchange.

More on this:
here
here
and here

That may be how traditional optical communications works. Quantum crypto, otoh, relies on the light being put in a certain polarization state by the sender. It's designed so that a stream of single photons go from sender to receiver; there can be no equipment in-between. If an intermediary views this photon en-route, it disturbs the polarization seen by the receiver. Because of the way the sender and receiver can agree on which photons were correctly measured, any aberrations (intercepted photons) are discarded. The most you can hope for is a denial-of-service.

Here's a better explanation than I can muster.

If you've got a bit of maths under your belt, or even a bit of coding would suffice, there is a link on this page to some Matlab code used to detect steggafied images.

Modern schemes wouldn't be necessary because quantum cryptography would become the standard and is proven to be unbreakable by the laws of quantum mechanics. Any interaction (malicious or otherwise) of a third party is noticable to the proper parties and the message/key transmission is just repeated until a clean send is achieved.
Here, here and google (of course) provide some good reading if you're interested

Unfortunately, this attempt is headed by a guy known for using dependent variables that are not well agreed on.

Unfortunately, this guy is also known for then using manipulations that do not respect the limits of his independent variables

That's why Herrnstein and Murray's previous "work" is termed ideology, not science.

We might hope that risky theses are attempted, to push the limits of what can be known, to enlighten and inform the debate, if not to provide definitive answers. No one's against that. But it's a fact of life that if your key staff include people known not to be equal to the task, your conclusions won't be taken seriously.

In short, Herrnstein and Murray gave up the right to be taken seriously in this field when they published their "work" last time. Perhaps some day a credible proponent of this thesis will come forward, who can be taken seriously.

Dartmouth just recently stopped billing students for long distance calls because administering billing was more expensive than the total charges... no doubt this VoIP initiative will help them save on total campus phone bills without overburdening the network. Dartmouth has been pretty chill about P2P-- they allowed a home-grown P2P program to be used for awhile. I'm pretty sure that even back in the heady days of Napster, they never totally banned its use, but rather metered Napster bandwidth. Regardless, they've recently upgraded network capacity to connect hubs with gigabit ethernet, and don't anticipate VoIP being a bandwidth burden.

To skip right to the heart of it, from the article, here's how it can be broken:

"Quantum cryptographic techniques provide no protection against the classic bucket brigade attack (also known as the ``man-in-the-middle attack''). In this scheme, an eavesdropper, E (``Eve'') is assumed to have the capacity to monitor the communications channel and insert and remove messages without inaccuracy or delay." http://www.cs.dartmouth.edu/~jford/crypto.html

Not exactly "unbreakable". Thanks to whoever posted that link earlier.

You can find more info on Tillman Grengross, the only scientist mentioned in the article, at his faculty homepage at Dartmouth.

Oh really? This has nothing to do with the Good Sam laws. Those are intended to prevent lawsuits from helping injured people. Misprison is about the Feds getting everyone related to a federal investigation. Don't get mad at me because it doesn't make sense. I didn't write the law. Here you go: Misprison of a felony

No need to be a jackass. It's among the craziest crimes on the books but it's real. I know somebody that went down for it. Here you go: Misprison of a felony

No it isn't. I know someone that went down for this among other charges. You need a reference? Here you go: Misprison of a felony

Here you go: Misprison of a felony

Quote: Machines infected with the worm are programmed to launch a denial of service (DoS) attack against Microsoft's Windows Update website on the 16th of each month, starting in August 2003.

vnunet site has been down for a little while... here are the appropriate google cache links:

Part 1
Part 2
Part 3

----------------
karma whoring for linux 'n stuff: Dartmouth Open Source Community

"In secret-key encryption, a k-bit ``secret key'' is shared by two users, who use it to transform plaintext inputs to an encoded cipher. . . A key of 128 bits used for encoding results in a key space of two to the 128th (or about ten to the 38th power). Assuming that brute force, along with some parallelism, is employed, the encrypted message should be safe: a billion computers doing a billion operations per second would require a trillion years to decrypt it. . .

"The main practical problem with secret-key encryption is determining a secret key. . . A possible solution is to agree on a key at the time of communication, but this is problematic: if a secure key hasn't been established, it is difficult to come up with one in a way that foils eavesdroppers. In the cryptography literature this is referred to as the key distribution problem. . .

"Quantum encryption provides a way of agreeing on a secret key . . ."

"In secret-key encryption, a k-bit ``secret key'' is shared by two users, who use it to transform plaintext inputs to an encoded cipher. . . A key of 128 bits used for encoding results in a key space of two to the 128th (or about ten to the 38th power). Assuming that brute force, along with some parallelism, is employed, the encrypted message should be safe: a billion computers doing a billion operations per second would require a trillion years to decrypt it. . .

"The main practical problem with secret-key encryption is determining a secret key. . . A possible solution is to agree on a key at the time of communication, but this is problematic: if a secure key hasn't been established, it is difficult to come up with one in a way that foils eavesdroppers. In the cryptography literature this is referred to as the key distribution problem. . .

"Quantum encryption provides a way of agreeing on a secret key . . ."

AFAIK, John Kemeny and Thomas Kurtz invented BASIC at my college a while ago and then gave it away for free.

Mr. Gates was just riding on someone's coattails, near as I can figure...

tsk, tsk, tsk . . . :-)

Ummm, a multi-demensional arrary of numbers was re-loaded and the Trinity, (Godhead, Father, Son) is now dead? I don't get it. Is this a reference to the following? ....

In the latest effort to convert the mathematically squeamish, Dartmouth College has developed a new program, "Math Across the Curriculum," that integrates the subject into some very unlikely places -- art and literature, for example.

"A Matter of Time" pairs a comparative literature professor and a math professor to lead discussions on the fourth dimension. Readings include "One Hundred Years of Solitude" by Gabriel Garcia Marquez, the Book of Genesis and works by Freud, Nietzsche, Poe and Borges.

There are a bunch of local ISPs in the area that might be willing to do this. Based right in Hanover is segNET, which provides a lot of the bandwidth to companies in the area. Across the river in Norwich, VT is ValleyNet, which was one of the first ISPs in the area (it now has something like 5,000 - 10,000 subscribers).

ValleyNet also happens to be a non-profit organization originally formed by Dartmouth and the Montshire Museum of Science. VN is always trying to help get the community connected to the Net so if they're something that they can contribute, they probably will.

Most of you probably don't realize just how backwards the Upper Valley is. Until recently DSL in the area was only offered by a couple of places and cost upwards of $150/mo., even for individuals. Thanks to ValleyNet people can now get low-speed DSL for about $30/mo.

And for you cable fans out there: there are *no* cable modems in the area. The local cable company is Adelphia, who (last I heard) suspended the installation of the new equipment needed for cable modems back when they filed for bankruptcy.

Why do I know all this? Because I live right across the river from Hanover and I used to work for ValleyNet.

There are statistical techniques for determining if they've been altered.

A language is the set of all ways a grammar allows symbols to be combined.

(of course, a grammar is a set of rules on how to combine symbols.)

Under the formal definition, XML is indeed a language. It is not a language useful for defining algorithms, admittedly.

Can we stop with this "XML is not a language" now?

Clean from the point of view of the person implementing the logging or DB transaction code, perhaps... but think about it from the point of view of the person debugging the system: at any method invocation in the original code, some other arbitrary code might run, with possibly arbitrary side effects, including possibly invoking other methods and/or changing instance variables in the object that's currently being used.

Of course, no sensible person would use this kind of feature in a way that would cause these kind of side-effects, but when debugging a system you don't know whether the original implementor was sensible. Perhaps they made a mistake (something I believe has been known to happen...).

The point is that by adding primitives like this, you've gone from building your system on solid building blocks to a foundation of shifting sands. An API that was stable before might be utterly demolished by a slightly injudicious use of this feature.

Those people in this thread comparing this feature to LISP macros were dead on... experienced LISP programmers use them seldom. As this says:

Macros are powerful, but dangerous. Use them sparingly, and only when you really understand what they might do.

Yeah, we've been warned for about 20 years now. And they've already solved the problem.

• biography
  
  
• a study guide for Zarathustra
  
  
• an ecncylopedic reference in German and English
  
  
• secondary sources, all online
  
  
• original writings (English translations) and links to criticism
  
  
• Hypernietzsche (very advanced)
  
  
• Nietzsche listservs

This is a PPT, but hits the main points: Challenger Disaster. An ugly page that has an actual paragraph is this. But I finally found a real page here.

Ha! You seem to be thinking of a different Dartmouth prof.

I have no idea what the parent post has to do with this article, but it's very much on-topic for me, because I worked for him a few years ago. That Dr. Strangelove scene came up a lot among his students... (For the record, Masters really knows his stuff; unfortunately, I don't think that biochem is his stuff.)

...but little progress. I'd say that the one area programming has made leaps of progress is in functionality. But if reusability is the measure of progress, and it is a requisite, then little progress has been made since McIlroy formally introduced the world to highly reusable software components.

If I may actually provide an opinion rather than just rant (I know, it's dangerous here at rant.slashdot.org)...

Design By Contract is an excellent step towards highly reusable components. In order for a component (in its loosest definition) to be reusable it must also be correct. But correct is relative. The requirements, or the contract, must be apparent to the user (client). The run-time should check the requirements for you, and this run-time checking should have the ability to be turned off. Furthermore, in the object-oriented case, the contract should be inherited providing a framework for more specified descendants. DbC is a good tool for providing these.

This is a .sig

If you want your users to love you, check out Dartmouth's
Anyone who has ever been to Dartmouth or any other school using a BlitzMail installation will vouch for the strength, ease of use, and plain usefullness of the system.