Slashdot Mirror

Anyone know what I would need to learn and do in order to get involved on a theoretical or practical level?

There exist effective techniques that can anonymize the data in order to thwart attempts to correlate identities, while still preserving the statistical properties of the data that make it useful to researchers. They include k-anonymity and l-diversity:

http://privacy.cs.cmu.edu/people/sweeney/kanonymity.html

http://www.cs.cornell.edu/~dkifer/papers/ldiversity.pdf

http://tera-3.ul.cs.cmu.edu/ULIBHelp.htm#faqBkMark

http://tera-3.ul.cs.cmu.edu/cgi-bin/udlcgi/ULIBSimpleSearch.cgi?title1=hitchhikers&search=Title+Search

They've got "HitchHikers Gljdie to the"..... is it too hard to even spell check the titles?

Actually, that's called a URL. A link would be this.

True, but in the real world, it's not as simple as that. There are cases of publicly available databases that you gave no permission to grant access to (for example, AOL's release of their search queries). There are other cases when a database has restricted access, but a person with access to it takes it and uses it in comparison with other databases available. Hackers are always a trouble; since some have gotten into such "secure" areas as the CIA and IRS, what's to keep them from potentially getting into any database?

The problem is one of privacy - in the worst case (or, for those who are cynical, common case) we have none. There's been some answers proposed to solve this. If you're interested, I'd start by reading the original paper on k-anonymity, which attempts to create privacy in a world where one can possibly have access to any database, ever. It can be found here: http://privacy.cs.cmu.edu/people/sweeney/kanonymity.html. (There are, of course, a multitude of other methods; k-anonymity is just a good starting point.)

Nice try, but a protocol for perfect anonymity has already been described:

http://www.ece.cmu.edu/~adrian/731-sp04/readings/dcnets.html

Not sure if there are any actual implementations yet.

Unfortunately I can't make much of your link -- most of it is in German, which I can't read, and there is a huge amount of material that would take me a decade to process. I have no idea where I'd even begin to make a comparison.

I have yet to see a specific example of a dead field that is not superceeded by some other field that answers the same questions in a better way, or made trivial by computers (both of these cases I would consider to be compression of knowledge). Please do point one out if you know of such an example.

There are a smallish number of modern books which teach it, but most staple DE courses have regressed to teaching collections of tricks and simple numerical methods, while courses on Lie algebras have other priorities and certainly don't teach any of that, even though it was the birth of the whole subject.

FWIW the in the example I cited the author discusses explicitly the goal of compressing proofs and explanations.

Wouldn't a second sensor in your mouse, or another wire to a chip in there be a fantastic way to provide entropy?

I don't really understand the focus on having anthropomorphic robots. Humans are relatively
bulky and require quite a bit of energy for locomotion and so far move fairly slow. The one
benefit of bipedal locomotion is the ability to walk over multileveled and rough terrain, but
I'm guessing the majority of robot uses will be in offices, homes, etc. Why not have more designs
like this? http://www.msl.ri.cmu.edu/projects/ballbot/

They do. Federal law-enforcement is always present at, and typically presents at, APWG meetings (I work for an APWG member), and they do track this stuff, and when possible, make arrests. Among the problems they face are volume (there's so much of this stuff, and LE does not have unlimited resources), time (doing the investigation and compiling evidence is by its nature very painstaking work), and the fact that the perps are most commonly in Russia and other eastern European countries, making apprehension and prosecution far more difficult.

They can't solve all the problems, or maybe even most of them, but they're doing what they can, and it's more than you'll read about on Slashdot. No matter how much resources the FBI and others throw at this problem, however, it will always remain mostly a problem of technology combined with user education.

At the last APWG meet, in Pittsburgh, some researchers fron Carnegie-Mellon presented there findings of an anti-phishing game they wrote, the idea being that you can more effectively train users to not be phished by having them play a video game, rather than read some boring instructions from the IT department or watch a similarly boring video. Their test subjects showed real improvement Vs. a control group, and there has been considerable interest in the game.

A preview version is here, for anyone interested:

http://cups.cs.cmu.edu/antiphishing_phil/

License is CC-attribution-non-commercial.

(I am not affiliated with CMU)

Skyrates is not really a time-killer, but it is easy to get heavily involved in it...

Someone doesn't know their history. Parent is incorrect.
http://www.cs.cmu.edu/~awb/linux.history.html

Cyrus includes Sieve if you want to filter on your server.

This project predates google's scanning project by several years. Brewster tried to get google involved, but as usual they decided to go alone. While the OCA was announced in 2005, it was an offshoot of the Internet Archive/CMU/Raj Reddy's Million Book project which was started in 2001 with books being scanned in India.

I think I may have started this back in 2001, when the CSS stuff was current http://www.cs.cmu.edu/~dst/DeCSS/Gallery/Stego/index.html
And yes it is VERY simple.

Even "genetic algorithms" are not evolution in the strict Darwinian sense, because people intentionally design algorithms and then let them compete.

Not always. For example, see psoup, Tierra, and Avida.

For this to be truly genetic, the algorithms themselves would have to evolve.

For weather simulations to be truly predictive, the algorithms would need to condense out of the atmosphere.

For traffic simulations to really work, they would need to pave little digital roads and build little digital cars.

For nuclear physics simulations to really be nuclear, they would have to require lead shielding around the computer.

Throw them out. If the voter doesn't have enough brains to fill a circle in front of a name then chances are he is too stupid to vote.

Except that the courts generally regard the right to vote as being fairly universal - there is a great reluctance to invalidate votes, hence all the argument over hanging chads. The whole Florida fiasco wouldn't have been a big deal if you just told the machine to count the ballots and defined a vote as whatever the machine could pick up - if you voted "wrong" then that's on you. Most people don't accept this, so the voting system has to handle it.

I don't need an overlord who forces me to vote for a local dog catcher. I don't care, and I want to abstain, it's my right.

Abstaining is perfectly fine - an abstained vote is NOT an ambiguous vote. A computer-generated ballot would allow voters to abstain, and when they do so it would unambiguously generate a ballot that indicated that they did so.

To that effect he is entitled to spoiling this section of the ballot (if not the entire ballot, which is also his right.)

If the voter wants to mail the electoral board a rant they are welcome to do so. However, the electoral system does not need to be engineered around allowing ballot vandalism. They can always write-in Mickey Mouse if they want to cause mischief. We don't let people shout and scream in court, and we don't need to allow people to write on ballots to the extent that it lowers the reliability of the voting system.

I'd like to see an example of "millions of machines that work without defects".

This would be a 6.25 sigma process. Airlines exceed this level quite reliably. This site suggests that most industries run at four sigma. Six-sigma processes are fairly common in the industrial world - just look at how many units most companies ship. I think your X-box figure is highly inflated, and even a 3% defect rate would probably not be tolerated by many retail outlets (that's a LOT of returns). In any case, we're talking about critical failures here - a spec of splattered paint on the printer case would not impact voting.

Don't get me wrong - I think that government is tremendously capable of messing just about anything up. However, for whatever reason they manage to run elections just fine most of the time (to the extent that they don't it is usually political in nature - and NOTHING will stop that).

While loaded with buzzwords, this really involves nothing that's really new.

Yeah, I haven't looked at it too intensively myself yet, but the impression I get is that most/all what Hawkins proposed has been proposed in the past. He basically took what was done in the past and made it much more accessible, which is great and all, but he really should've cited more of the prior work by others (or been more aware of it). Besides Grossberg, I think there's also quite a bit of similarity with the work of Rao & Ballard (1999) and Lee & Mumford (2003).

Still, I credit Hawkins quite a bit for making the general public much more aware of this sort of modeling.

Not that it would be entirely on topic but you just remembered me of this guy who took high resolution pictures of the ISS, the Space shuttle, some spy satellite, and Mercury. He took multiple frames of the object of interest and selected the best for combining well here it is what he exactly did:

"Images of Mercury were obtained at 8 bits of resolution using exposure times of 16.7 ms at a rate of 60 frames per second and recorded on broadcast-quality videotape for subsequent data reduction. The images were sorted and selected based on maximum gradient of the planet's bright limb, co-aligned, and added in 16 bit space."

Here is a link:

http://www.journals.uchicago.edu/AJ/journal/issues/v119n5/990240/990240.html

let me store it here so I don't forget it ;)

This method is used to reduce seeing which is random and capable of reducing that Mt. Wilson telescope aperture from 1.5m to some seeing limited aperture of maybe 10 to 20cm (my guess).

The pixelation problem is not that similar to seeing. I would think since each new larger pixel is the average of the same region in the original image it is somehow low pass filtered spatially. Recovering a part of the image would somehow make it necessary that the image information is stored in the time domain since you can't get it from that 2d single frame space. That high level view makes it again look like the seeing problem but with the seeing one gets some images which cover a large spatial bandwidth but have low dynamic range, while with the pixelation the spatial bandwidth is constantly low but the dynamic range in the image is constantly high.

Searching google gives me something about de-identification and how the simple methods here discussed are easily thwarted by face recognition software. I.e. the bad guy crosses the US border is photographed, later produces de-identified compromising images of himself. Then the blurred/pixelated image is fed into the face recognition program and compared to the border database - success should easily follow, because the facial features are still recognizable to the software.

Here is an example: http://reports-archive.adm.cs.cmu.edu/anon/2003/CMU-CS-03-119.pdf

I just can't find what you were talking about, and I'll come across who knows what if I try to find it myself:

http://www.cl.cam.ac.uk/~mgk25/ieee02-optical.pdf

Well, I did a quick google search, and here's what I found. It would appear that cell phones and other wireless devices can cause issues with the GPS systems used in planes, which are critical for landing the aircraft. Now, I don't know about anyone else, but when I get on a plane I want to know that when I get off the plane, it won't be in a farmer's field or an ocean.

That's all the same guy (in collaboration with various others). So no wonder you picked up the connection. :) http://www.cs.cmu.edu/~biglou/research.html

Perhaps time management isn't that important in the end, or perhaps the limited amount of time each of us may have makes it even more important.

Yeah, that's the question I had when I read through his PowerPoint slides yesterday morning, after the WSJ video came up in the course of my daily hour of mindless Fark surfing.

Pausch's methods are great for people who value a highly-regimented life, or who require the same to accomplish anything at all. There are people like that, and maybe he's one of them, but he overgeneralizes to a criminal extent, IMHO. Most of the worthwhile things I've accomplished can trace their beginnings to sitting around daydreaming and doing not much of anything, or looking for an excuse to put something else off. Hell, I wouldn't have seen his video and slides in the first place if I hadn't been killing time surfing the Web, right?

Ultimately, I spent half an hour watching the slides, and then went back to finish my daily list of unimportant links on Fark. I'll admit I was a little unsettled by one of the suggestions he raises ("Write your own eulogy. What do you want it to say?") because frankly, I don't know if people would find enough interesting things about me to even bother attending my funeral, and of course that bugs me. Everybody wants to leave a meaningful legacy, right? But ultimately, living by what you want your eulogist to say is just another way of living according to other people's standards. I finally managed to refute Pausch's dictums by imagining myself trying to persuade a Zen Buddhist practitioner to follow them.

You have to go your own way in life, and if you're lucky, you'll have the chance to determine how you die as well. If I were in Prof. Pausch's shoes, I'd like to think I'd wrap things up on my own terms, with a .38 Mannlicher and a one-way ticket to Washington, D.C. Consequently, my eulogy would depend entirely on how I behaved during the last five minutes of my life. Why should anyone be all that concerned about what people have to say about my Fark and Slashdot habits?

>I need to read and audit the source and then compile that source to be completely sure!

Even if you read the source, and compile it yourself, you would not be 100% sure it is safe.

Reflections on Trusting Trust ;-)

Theo de Raat, even though I don't agree with all he says, is a Free Software lover, and I don't believe for one minute that he would defend using a proprietary compiler for OpenBSD, an operating system designed for security.

If you *have* to wonder why, check Tusting Trust

http://www.cs.cmu.edu/~sef/Orig-Smiley.htm

---------------

Original Bboard Thread in which :-) was proposed

Here is the original message posted by Scott Fahlman on 19 September, 1982:

19-Sep-82 11:44 Scott E Fahlman :-)

From: Scott E Fahlman

I propose that the following character sequence for joke markers: :-)

Read it sideways. Actually, it is probably more economical to mark

things that are NOT jokes, given current trends. For this, use :-(

The entire thread is reproduced below. We didn't have formal newsgroup threads in those days, but these are all the messages that mention the need for a joke marker or that use the :-) symbol.

This was retrieved from the spice vax oct-82 backup tape by Jeff Baird on September 10, 2002. The period covered is 16 September 1982 through 21 October 1982.

Credits: Here is the account by Mike Jones describing how this ancient post was retrieved. It's an impressive piece of digital archeology, with many contributors. I am grateful to Mike, to Jeff Baird, and to all the others who played a role in this effort. It is great that we can view this bit of Internet history once again.

Many people were involved in this computing archaeology success story. I (Mike Jones) kicked off the effort in February 2002 by looking through some old bboard program (Bags) sources, figuring out the filename that the post would likely be found under (/usr/cmu/lib/bb/general.bb), and asking Howard Wactlar, the former CMU SCS facilities director, whether the file could still be restored. Scott Fahlman provided data narrowing the probable span of time during which the post was made. Howard and Bob Cosgrove, the current director, determined that backup tapes from that period (1981-1983) still existed and asked Jeff Baird of the facilities staff to try to find and restore the post. Dave Livingston of facilities located a working 9- track tape drive and a machine to use it on. Kirk Berthold and Michael Riley in CS operations managed retrieving tapes from off-site archival storage. Grad student Dan Pelleg's FreeBSD machine was used to read the 4.1BSD dump format tapes using a compatibility mode in the restore program. (Later in the effort a NetBSD machine was used to do the same thing.) Dale Moore looked for the post on Tops-20 backup tapes from CMU-20C. But by all accounts, Jeff Baird should get most of the credit for doing the hard work of locating and retrieving the data. He kept asking for more tapes, reading those that could still be read, narrowing the date range, and sticking with it until the post was found. Thanks all for your efforts to restore this part of computing history, and especially, thanks Jeff!

Note: There apparently were a few posts prior to 16 September (not on the tape that was retrieved) that posed various physics questions about what would happen to various objects in an elevator if you cut the cable. Given the quality of the elevators in Wean Hall (then and now), this was more than idle speculation.

Apparently someone had posed the problem of what would happen to a helium balloon in free-fall, someone else had asked about pigeons flying around in the falling elevator, and someone had then asked what would happen if the birds were breathing the helium...

16-Sep-82 11:51 James Wright at CMU-780D Related question

Of equal interest is how the birds cheeping will

sound after they have inhaled the Helium.

=

16-Sep-82 12:09 Neil Swartz at CMU-750R Pigeon type question

This question does not involve pigeons, but is similar:

There is a lit candle in an elevator mounted on a bracket attached to

the middle of one wall (say, 2" from the wall). A drop of mercury

is on the floor. The cable snaps and the elevator falls.

What happens to the candle and the mercury?

Most people have no idea that right exists.

I certainly didn't. Here's a DIY.

Only if you can also trust the compiler chain.

Slightly off-topic, but tangentially related to TFA: I'm in the process of writing my masters. I'm doing it on the NAS Conjugate Gradient (CG) benchmark to several exotic architecture. Now for those of you who haven't heard of CG, it's a very-commonly-used but extremely complicated algorithm. I wanted to have a section in my masters explaining how CG works, only I hit a snag - all of the explanations SUCK. I mean, REALLY SUCK.

I went to one of the profs in my department. He does numerical electromagnetism, so he is very good at math and CG is familiar to him. I asked him if he could recommend a "CG for dummies" book.

He told me, as a matter of fact, there is: An Introduction to the Conjugate Gradient Method Without the Agonizing Pain by Carnegie Mellon professor Jonathan Richard Shewchuk. My E&M prof said it was the best bit of technical writing he'd ever seen. I'm about halfway through, but I have to agree - though it's complicated, it's by far the most comprehensible explanation I have ever seen. It really is a perfect example of what technical writing should be like.

Can someone give me an introduction/explanation to ... functional programming that I can understand?

Scientology is so bizarre that I can't tell if you're being facetious or not.

Not only is this "church" legal in the US, the sheriff of the largest county in the state is a Scientologist. Sheriff Baca is able to get county money diverted.

And he issues CCW permits to prominent gun-banning Scientologists such as Sylvester Stallone. Wonderful.

On a more serious level, but for the same reason, there is no reason to think that this entry in the password file corresponds to a valid Unix password, since if that system was based on his code, he login will bypass normal authentication completely.

If you don't understand what I'm talking about - have a look at the paper. It's a classic, and well worth a read. Wikipedia has a summary .

Humphrey has written quite a few books on software engineering and management, and you would do well to read the lot of them. However, most software engineers would probably find his methodology too formal, as most programmers I know prefer a loose, casual working style. But there's no doubt that the SEI's methods are worthwhile when quality and reliability is important.

Also, before buying any technical book, check to see if the Associaton of C and C++ Users has reviewed it - try typing "management" into the search for their. (I'm afraid their book review server is down just now but I expect it will be back up soon.)

The ACCU makes a point of reviewing books that they suspect might be stinkers, so they can write reviews that warn you away from books that suck.

Humphrey has written quite a few books on software engineering and management, and you would do well to read the lot of them. However, most software engineers would probably find his methodology too formal, as most programmers I know prefer a loose, casual working style. But there's no doubt that the SEI's methods are worthwhile when quality and reliability is important.

Also, before buying any technical book, check to see if the Associaton of C and C++ Users has reviewed it - try typing "management" into the search for their. (I'm afraid their book review server is down just now but I expect it will be back up soon.)

The ACCU makes a point of reviewing books that they suspect might be stinkers, so they can write reviews that warn you away from books that suck.

Humphrey has written quite a few books on software engineering and management, and you would do well to read the lot of them. However, most software engineers would probably find his methodology too formal, as most programmers I know prefer a loose, casual working style. But there's no doubt that the SEI's methods are worthwhile when quality and reliability is important.

Also, before buying any technical book, check to see if the Associaton of C and C++ Users has reviewed it - try typing "management" into the search for their. (I'm afraid their book review server is down just now but I expect it will be back up soon.)

The ACCU makes a point of reviewing books that they suspect might be stinkers, so they can write reviews that warn you away from books that suck.

There have been some efforts in the area of networked filesystems with disconnected operations. I remember checking out AFS, Coda, and InterMezzo years ago. At the time, I found something wrong with each of them, but they may have improved since then. Of the three, I think Coda is your best bet.

Have a look at http://coda.cs.cmu.edu/ This is a disconnectable file system. It could be what you are looking for. Certainly, that is what I use for doing the same thing.

http://www.coda.cs.cmu.edu/

I'm slightly surprised noone has suggested Coda http://www.coda.cs.cmu.edu/; mind you I haven't read all the rest of the comments - this is Slashdot (& I'm at work)! It looks like Coda would be ideal for this sort of disconnected use of files from a file server, and I'm quite looking forward to giving it a spin on the next laptop I buy. Has anyone else got much experiance of using Coda? Does it actually work well, or is the synchronization a PITA, and what about its speed compared with that of NFS or SMB? Thanks for any feedback.

If you want to see a derivation, it was on my QM3 mid-term last fall... hardest class I've ever taken. The force per unit area between two large flat plates falls off as 1/R^4.

http://www.andrew.cmu.edu/user/bsauerwi/Problems/2 006QUA3MIDb.pdf -- problem #3. :)

~Ben

We found no substantiation for the "common knowledge" that
early Mach 3.0-based Unix single-server implementations achieved
a performance penalty of only 10% compared to bare Unix on the
same hardware. For newer hardware, [9] reports penalties of about
50%.

If The Open Group is "making standards work" (TM), then who is Making Work Standard?

If problems with pasting, maybe a link would do. And this one seems to cover the first use of plain chars smiley, in 1982. Tho in other system there were tricks to overlay chars and create single "cell" smileys, in the 70s. I also found a huge list of smileys.

They have a plan.

Here's a couple of articles I read:

Facts about the brain
Rods and Cones

There are around 125 millions rods and 6 million cones in each eye, with the percentages of each color/wavelength (red = 64%, green=32%, blue=2%)

No Sense

The human eye has 100 million neurons per per eye of five types, but there are only around 1 million neurons per optic nerve (arranged in bundles of 1000).

Let the user become the crawler- and do not eliminate the search giants (just don't rely on them completely). Already I sort of operate like a (slow) crawler with my queues of links to read, bookmarks (be weary- big load) and indexing those very interesting or important pages, sharing related tidbits, etc. Just feels like the natural extension, though I am sure that many people will want to stick with traditional GUIs and "back/forward" habits. There is also some interesting discussion in ATLAS-L re: future search infrastructures. So, in the spirit of promoting development in this area, linkage:

* Grub article (now defunct)- was distributed peer-to-peer crawler. (see also)
* Boitho, another distributed crawler
* YaCy- another peer-to-peer crawler
* How to build a web spider
* C++ web crawler lib
* LibWWW (perl)
* W3C's WebBot
* The Internet Archive's Heritrix crawler
* WebSPHINX- customizable crawler

Somehow, this is like an extension of surfraw. I imagine that soon enough we will start up an open source crawler-browsing hybrid software package, though have been surprised that nothing like it has popped up yet- it's (usually) the way of the programmer to make sure that he has the ability to do what the giants are doing. Maybe we have all been collectively blinded by graphical web browsers (IE, Firefox, Opera, etc.) and "click-click-click" thinkware?

there is a video on google's tech talks by Luis von Ahn in it he talks about this and shows some of his on work in this field.