Slashdot Mirror

I'm not really karmawhoring, because I hit the Karma Kap long ago, I'm just linking to the Warhol Worm... :-)

If you're iteratively processing stuff (transcoding, editing, mixing, etc) it's way better to do the lossy step only once than each time. Just 'cause 1 iteration may be unaudible doesn't mean 5 will be. (indeed, transcoding is one of the best ways to bring out compression artifacts!)

I'd like to use a lossless audio codec for video capture and editing for this reason. Unfortunately, it doesn't appear that any of these lossless compressors are set up so they'll work as audio codecs inside video editors. I can use Huffyuv to compress video without loss; why aren't FLAC and its competitors available as ACM codecs so that content compressed with them can get stuffed inside a WAV or AVI?

Until then, I guess I'll just have to stick with uncompressed audio...when 1 hour of Huffyuv-compressed D1 video takes 20-25GB, the 605MB that one hour of CD-quality audio takes isn't much by comparison.

The United Devices (UD) Cancer Research project allows people to crunch data (much like SETI@Home) but instead of finding alien life the idea is to find a cure for cancer. The software (as far as i can tell) models how various chemicals interact (IANAChemist, so I can't really give much more detail than that - check out the site if you're interested of course).

Though this post to clarify the previous one may be getting offtopic :p

The suggestion is that dinosaurs went through a gliding stage before learning to fly with two fore-limbs, says Dr Angela Milner of London's Natural History Museum.

This sort of quote assumes that evolution is going in a single direction -- "from" flightless dinos "to" modern birds. In fact traits commonly appear, and disappear, and reappear, many times. (Take a look at a "terror bird" and convince yourself birds weren't turning back into dinosaurs.)

It sounds like the world had a mess of different uses for feathers, once they developed -- insulatory, locomotion, display, and so on, just like in modern birds, and some we haven't thought of like this four-legged gliding model, if the fossil's real. Dinosaurs didn't develop "toward" flight, they bounced all over that range of feather uses just like birds do today.

Cladistics will air out that sort of thinking real fast. (Decent practical primer/pop science book: "In Search of Deep Time.")

Looking at things in "clades" also helps in practical ways by showing the evolutionary relationships between living animals more clearly. People trying to figure out ways to treat tapeworms had trouble making progress under the assumption that their on parasitism evolved only once, in a common anscestor of all modern tapeworms. Cladistics hashed out the evolutionary history of tapeworms a bit, and we realized the trait had a more patchy history -- parasitism had evolved several separate times -- and that some of the closest modern relations weren't parasitic at all. Those modern relations were easier to work with in the lab than something that required a host.

CQual

It's been used to find security holes.

i'm a graduate student at the university of california, in the school of information management and systems. the ban on segways was proposed back in august, and we wrote a paper on the future of its use. i have it up in pdf format -- go ahead and have a look -- if there's any demand whatsoever, i'll convert it to html. enjoy.

seti account

must prove something

Ace's Hardware needs to research real servers before talking about their "scalable" servers. Their numbers are really saying that their box performs like a dog.

For those of you interested in this topic here is a few pointers and words of wisdom.

Server scalabilty and performance has three basic metrics, thruput (urls/sec), simultaneous connections, and performance while overloaded. Of course, you could add latensy but I'd argue that with the correct design latency is directly proportional to the real work you are doing, bad design insertes arbitrary waits.

I know of a HTTP Proxy by a large ISP that does user authentications & URL authorization (re: database), header manipulation, and on-the-fly text compression at 3000 urls/sec for 2000-4000 simultaneous connections and maintains that performance under load by sheding connections, all this on a dual 1GHz Intel PIII box running a Open Source OS that starts with "L". That is a maximum of 260 Million URL/day, three orders of magnitude greater performance than Ace's Hardware stats.

The simple answer to the question "How do I create a scalable fast network server?" is Event-driven GOOD & Threads BAD. Event driven network communication is two to three orders of magnitude better performing than thread/thread-pool based network communications. See Dan Kegel's C10K web page. That means you must use non-blocking IO to client sockets and databases. Once you accomplish that small feat, dynamic content just consumes CPU; with 2.8 Ghz Xeon processors you have plenty of cycles for parsing HTML markup or whatever. Threads cause cache thrashing, and context switching. While thread programmers don't see the cost in their code, just read the kernel code and you'll see how much work HAS TO BE DONE to switch threads. Event driven programming just takes some state lookups (array manipulation) and a callback (push some pointers onto the stack and jump to a function pointer).

Desgin is FAR MORE IMPORTANT than which runtime you use (execution tree, byte code, or straight assembly). I have done some very high load network programming with Perl using POE.

Python has Twisted Python

Java has the java.nio and the brilliant event/thread hybrid library SEDA by Matt Welsch.

I am also looking into the programming language Erlang which builds concurrancy and event driven programming into the language. Further, Erlang is used by some big telco manufacturers to great effect (high performance and claimed 99.9999999% nine-nines reliability on a big app).

Queuing approaches have proven to be much more scalable in other areas - no reason to think it wouldn't work for web servers. Check out SEDA: An Architecture for Highly Concurrent Server Applications for a working implementation in Java that outperformed Apache [insert benchmark caveat here].

More on event-driven servers that minimize data copies and context-switching here.

The collections include a tiny fraction of the culture we create. For example, it's estimated that there are 80 billion new images each year.

I thought Berkeley put IP in Unix... Am I mistaken?

If everything started with a Big Bang from a singularity roughly 10-20 billion years ago, how is it that things came to be physcially 13 billion or more light years apart? I understand that the "Doppler" redshift is caused by great speed away from us. But is the universe seriously expanding at anything near lightspeed?

Some suggest that initial expansion was faster than light speed, and that the Hubble expansion is accelerating.

for some really cool ``sound animation'' presented at SIGGRAPH this year, check out James O'Brien's research

this is not exactly what you're talking about, it's computationally generated sound from 3-d animation, which is much cooler...and harder.

Guys, instead of trying to do something whose only purpose is to allow people to rip off games, why not do something noble that will help humanity. Here are some suggestions:

1.) Seti@Home
2.) Cure Cancer
3.) Evolution@Home
4.) Entropia
5.) eOn
6.) Climate Prediction
7.) Particle Accelerator Design
8.) Analytical Spectroscopy Research Group

See a complete list here: http://www.aspenleaf.com/distributed/distrib-proje cts.html

And no, I don't consider cracking encryption "noble". Especially when people don't seem to get the point that if it takes tens of thousands of computers months and months to crack some encryption, it is GOOD ENCRYPTION.

Currently, Web content providers and have no mojo to abuse in the first place, which is no better for us all in the long run than the appearance and domination of the next RIAA-like organization. Either way, we, the consumers of content, risk losing out on some good stuff.

I disagree. It is better without, because said "RIAA-like organization" can't lobby Congress to limit/remove our freedoms in order to fatten their bottom line. That is, the point of business, after all. And if you don't believe me, you can believe this.

Honestly, I don't think a micropayment solution will arise until the Government insitutes some sort of official e-cash solution. Given that the general public is a horde of moronic technophobes, and the country is currently being run by one, I seriously doubt such a solution being implemented in my lifetime.

So, until then, web publishers can run their sites as ad-supported (or referral supported), or find a line of work that will actually pay them actual money, and stop bitching. Nobody's forcing them to run a website.

1. Provided Microsoft uses a proper public key infrastructure, brute-forcing this thing could potentially take forever

2. This so that you can feel good subverting an X-Box by making it run Linux

3. By that time the hardware would be definitely obsolete, or X-Box 2 would be out with programs signed with a different key

4. And in any case, buying the X-Box already helps Microsoft. The more units sold, the more games developed.

5. There are tons of other worthwhile distributed computing projects to do out there - Folding@Home, SETI@Home, Mersenne Prime Search etc.

Grow up folks! Running Linux on a hacked X-Box is cool, yes, but this might be going too far...

Here
And Here

My favourite is personally: This beautiful piece of art.

Here
And Here

My favourite is personally: This beautiful piece of art.

Here's a short listing of some I know about (thanks to another poster for reminding me about KDVS):

• Pittsburg: WRCT
• Georgia: WREK
• New York: WFUV
• Los Altos, CA: KFJC
• Berkeley, CA: KALX
• San Francisco, CA KUSF
• Davis, CA: KDVS
• Stanford, CA: KZSU

One thing to remember about college radio is that it changes constantly from program to program... there's often very litte of an attempt at presenting a consistent sound as in commercial radio. So don't just listen once or twice to a station and assume you know what they're about. Maybe you should look for an online program schedule to figure out when to listen.

KALX, UC Berkeley's station, streams if you aren't within a mile or two of campus (90.7 FM for locals). It's easy to find some DJs/shows that focus on your preferred genres, although I look forward to new experiences. Good times.

KALX, UC Berkeley's station, streams if you aren't within a mile or two of campus (90.7 FM for locals). It's easy to find some DJs/shows that focus on your preferred genres, although I look forward to new experiences. Good times.

In fact if entropy did always increase you could use that to build a perpetual motion machine. If it did always increase you could make predictions about particles that could be exploited by a Maxwell type demon. But the fact that it usually increases, but might sometimes decrease, means that information isn't available to you. (Similarly if there really were such a things as a "law of averages" people could use it to win at roulette. That hasn't stopped people trying though.)

Some people have already produced excellent results in breaking visual CAPTCHAs.

Yes, it's possible, and has been done recently by some guys in CS at Berkeley. Breaking captchas had always been posed as an open challenge to the AI/image processing community.

NY Times article

Berkeley press release

Computer vision pages (w/papers)

Greg's page on breaking Gimpy

Yes, it's possible, and has been done recently by some guys in CS at Berkeley. Breaking captchas had always been posed as an open challenge to the AI/image processing community.

NY Times article

Berkeley press release

Computer vision pages (w/papers)

Greg's page on breaking Gimpy

Yes, it's possible, and has been done recently by some guys in CS at Berkeley. Breaking captchas had always been posed as an open challenge to the AI/image processing community.

NY Times article

Berkeley press release

Computer vision pages (w/papers)

Greg's page on breaking Gimpy

Personally, I won't be bothering with eBooks until I have a 300DPI+ A4-sized display*, but anyway:

Of course "readable to the human eye" is not the same as "pleasantly readable to the human eye" - you could just buy a print book if the ebook becomes too annoyingly full of wobbly characters.

And also, researchers, spurred on by the challenge of descrambling those obfuscated text things, are already having some success. See

"Breaking gimpy: Researchers crack Security System Designed to prevent internet Robots"

* LCD Manufacturers: I want a high DPI screen, not a physically huge one. Why the hell can't I get a 15" 1600x1200 DESKTOP LCD Monitor???

Something which destroys its own host hampers its ability to spread

That's not really a problem for computer viruses. They can scramble the harddrive and keep infecting other computers. Just read the addressbook first to send out an E-mail infections. You don't need a harddrive to scan the internet for other machines to infect. Wipe the data and leave the machine running.

Just as signifigantly, a properly designed virus could cover the entire internet address range within hours, possibly even in mere minutes.

-

Likewise, every remote root exploit makes it technically possible for this to happen. Even if relatively few people are being hacked by script kiddies today, that says nothing about the odds of a highly skilled attacker pulling off a single massively devestating attack.

This report is no reason for complacency.

I see why they're doing what they're doing. It may sound trivial to you to revamp a lesson plan, but

1) The point of having an experienced professor is that they can *reuse* what they've learned for previous years. I had an excellent professor, Prof. Rudich, at Carnegie Mellon University. He put it pretty well -- "I've found what I consider the most important, the most enlightening problems I could turn up over the years. I could change them from year to year, but then I wouldn't be giving everyone the *best* questions I've found." And his assignments, while decidedly tough, really were incredibly good.

2) It may sound really minor to you to swap in new programming assignments. However, it really isn't. Designing and debugging a lab is a lot of work. I've found that frequently labs aren't really solid and clear until their third year or so in use.

3) It's really frusterating to professors who teach classes that might have this one as a prereq to have half the class have covered a particular issue (because one lab brought it up) and the other half not.

I think a better solution, though perhaps a bit harder to enforce, is to allow students that have already written code to do whatever they want to with it. However, a copy of their code *must* stay on file.

Then, professors can use moss or something similar to look for duplicate code.

I feel that this approach, which would let students do what they want with their own code, would also provide a more realistic simulation of a production environment. When you're writing commercial code, there's always plenty of GPLed code out there to take. But if your license conflicts with the GPL (or your contract states that you need to write everything yourself for some reason, which happens here), and you run out and break it and then get caught, you get hung out to dry.

Attempting to stop the problem by preventing sharing of answers is difficult to prevent (if it's done privately), difficult to enforce (what if the person has graduated or is at another university that had a very similar assignment?).

Finally, I want to mention that I support at least some effort to avoid cheating. Simply "trusting students" sounds wonderful, but utterly fails in practice...the level of cheating in CS courses astounds me sometimes. Yes, certain types of assignments are more cheat-proof, but are also less useful in teaching.

Due to a flaw in the implementation of WEP, 128-bit encryption ends up being no more secure than 40-bit. More info here.

Sadly, the problem is not that there is no encryption in the hardware, but that the encryption built-in to 802.11 hardware (WEP) was not implemented properly. Here is a nice summary of the problems. Essentially, WEP can be cracked in a very short time with minimal effort. Using something like IPSec or SSH to encrypt the data in software before it is transmitted can help secure the data, but does not prevent someone from using the network for other purposes. In the case of businesses or government offices, if the data needs to be kept safe, it should remain on a wired network, with no connection to the outside world.

I think that has always been the MS line. Sometimes it is stated as: ``You folks here at _____ can't possibly justify the expense of developing what our [lying] salesmen [wrongly] claim that we can provide.''

Some of the best minds in the world have been working on open source and libre software [1] since before MS first stole time on a university mainframe. They've always worked to make systems which filled their needs, and so open source and libre software has always been aimed at what we used to call power users. It has been hard to learn, easy to use, and powerful.

MS has succeeded in large part because of their marketing efforts. Opensource/libre has succeeded as far as it has because of technical merit. Now we see some marketing being applied to these same systems which were making it on their technical merits. Redhat, Mandrake, and now Debian desktop are all making libre software accessible to the clueless. FUD seems to be loosing its effect. MS is starting to panic. Hence, we see this sort of ill-thought-out nonsense from the salesmen.

1. Think of Knuth and TeX, or Fateman, Schelter, et al and Macsyma/Maxima.

Yes, I mean, if they're going to do something like that, I could just say "i'll do it" and run something useful, like Seti@Home in its place.

The sad part is, I know at least 20 people that would keep it on for ambiance in their home.

well now that we have courses at major universities all about weblogging we may have a seperate part of our job description set aside just for weblogging!

Microsoft LOVES patents. They even love paying for them AT A VERY high price. It helps exclude the competition.

The outcome is already known, Microsoft will capitulate with InterTrust and will make it cost huge bucks. InterTrust then gets to yell "see, my patents stand up to scrutiny and they are really valuable" and nukes everyone else out of the DRM buisness by charging over the top for licencing fees.

Fortunately for us, DRM is a flawed concept and anyone who ever tries to get to close to it ends up paying with lost customers and diminsihed loyalty.

A paper by Carl Shapiro discusses this argument better than my rant.

BTW - this practice is not exclusive to Microsoft.

I just finished a CS course co-taught by Professor Patterson, and our primary text this semester was Patterson and Hennessy's Computer Organization and Design.

When we discussed pipelining this semester, the analogy used was the four stages of doing laundry: washing, drying, folding, and stashing. Here are the lecture notes (both PDF). The notes spend a good deal of time going over the hazards of pipelines and how to avoid them.

I just finished a CS course co-taught by Professor Patterson, and our primary text this semester was Patterson and Hennessy's Computer Organization and Design.

When we discussed pipelining this semester, the analogy used was the four stages of doing laundry: washing, drying, folding, and stashing. Here are the lecture notes (both PDF). The notes spend a good deal of time going over the hazards of pipelines and how to avoid them.

I just finished a CS course co-taught by Professor Patterson, and our primary text this semester was Patterson and Hennessy's Computer Organization and Design.

When we discussed pipelining this semester, the analogy used was the four stages of doing laundry: washing, drying, folding, and stashing. Here are the lecture notes (both PDF). The notes spend a good deal of time going over the hazards of pipelines and how to avoid them.

I just finished a CS course co-taught by Professor Patterson, and our primary text this semester was Patterson and Hennessy's Computer Organization and Design.

When we discussed pipelining this semester, the analogy used was the four stages of doing laundry: washing, drying, folding, and stashing. Here are the lecture notes (both PDF). The notes spend a good deal of time going over the hazards of pipelines and how to avoid them.

• The CHIPS satellite is a, well, satellite. The CHIPS home page clearly states that it will "be launched into a 94 inclination, 600 km circular orbit."
• Just because it's called an "Interstellar Plasma Spectrometer" doesn't mean it's on an interstellar mission.
• The mission will last one year (see http://chips.ssl.berkeley.edu again). It took Voyager about a year and a half just to get to Jupiter. Thus, this thing could not get "interstellar" even if it tried.
• Light takes 8 minutes to reach Earth from the Sun (Sol).
• Light takes about 4 years to reach Earth from the nearest star other than Sol (Proxima Centauri.)
• Radio waves are light waves and are thus not much much slower. By definition, light (and thus radio) travels at the speed of, um, light.
• Since the bloody thing is going to be in LEO (low earth orbit,) they won't have to communicate with the machine the way you describe.

Here are some website links:
Spacedev is (duh)
www.spacedev.com
CHIPSat can be found under Missions.
Spacedev's stock price can be found at:
finance.yahoo.com
The Space Sciences Laboratory at UC Berkeley
chips.ssl.berkeley.edu

All somebody has to do is post a ftp://chips.nasa.gov/pub
link.

It may become the orbital equivalant of goatse.cx, plagueing scientists forever.

A picture of DSS 62: The dish that picked up Pioneer 10
http://www.mpifr-bonn.mpg.de/div/vlbicor/pic_htm/d ss62.htm

PIONEER 10 AT ARECIBO
http://www.seti.org/science/ao-p10.html

Pioneer Home page
http://spaceprojects.arc.nasa.gov/Space_Projects/p ioneer/PNStat.html

Earth (the dot in the middle) as seen from 3.7 billion miles away by the Voyager 1 spacecraft, on 6/6/1990:
http://setiathome.ssl.berkeley.edu/pale_blue_dot.h tml

A Ride Under the Arecibo Radio Telescope
http://www.seti-inst.edu/science/under_the_mesh.ht ml

Merkac Dot : 48210

Links to Google Cache(N.B. Not always cached.)

article cache [Link not cached at time of posting]
Public Library of Science(PLoS) cache [Cache link active]
Nobel cache [Cache link active]
Harold Varmus cache [Cache link active]
Pat Brown cache [Cache link active]
Michael Eisen cache [Cache link active]
journals cache [Cache link active]
journals cache [Link not cached at time of posting]
Gordon and Betty Moore Foundation cache [Cache link active]
Licence cache [Link not cached at time of posting]
editorial board cache [Link not cached at time of posting]

Thank you kindly for your assessment. From your other postings it appears that you work at the University of Chicago... And I notice some references to the scavenger hunt, as well...

FYI, I used to work at NSIT, supporting the USITE clusters. You may have been over to the USITE:Crerar lab. We set up all of the visualization and collaborative computing over there. Anyhow, I designed the basic builds of those workstations, and hacked the Windows NT registry to run the SETI project as a low level background daemon (Service), and not just as a screensaver.

You should talk to Scott Wilson about hacking the USITE build, and replacing the Seti@Home daemon with a daemon version of your Sloan survey data viewer. It would be sad to see my USITE:Crerar SETI record go away, but I'm working on other stuff now days. Nab the USITE grid, and hack the registry to run your precompiled Sloan survey data binaries as a low level process. Doing so would be a good way to test the resource negotiation across a distributed computing environment (i.e. The USITE cluster... you would have at least three sites: Harper, Crerar, and Regenstein, running a distributed test environment for a larger scale (possibly global?) distribution of the Sloan survey data viewer)

Postscript: Where are you located? LASR? GeoSci? Argonne?

post-postscript: Check out Dr. Wimsatt's colloquium and lectures on Big Problems, Cultural Evolution, and the Dimensions of Globalization if you get the chance. I'm sure he and Dr. Foster are going to collaborate some day and bust up all sorts of new stuff for grid networks... Wimsatt's got all of the grid population genetics and memetic theory stuff worked out.

I trust a review if the reviewer seems knowledgeable and insightful. I buy a lot of opera on DVD and it's pretty easy to figure out who knows what he's talking about and who doesn't. With cookbooks, I look for people's description of what actually cooking the dishes was like. With technical books, I skip the "I loved it, you should buy it too" reviews and head for the long ones that discuss in depth the strengths and weaknesses of the author's presentation of the material. With other topics, YMMV, but this has worked for me and I have generally felt that I understood pretty much what I was buying.

Consumer reports it ain't, but most of what I buy at Amazon falls into the category of experience rather than search goods (see G.J. Stigler, "The Economics of Information," J. Pol. Econ., 69, 221 (1961); see also H.R. Varian, "Economics and Search," Plenary address at ACM SIGIR 1999). The question is, of what value is the time it would take you to research the quality of Amazon merchandise via a more trusted source than customer reviews?

This thing at Berkeley is smarter than you.

Berkeley says they've broken ez-Gimpy

I don't understand why companies don't include such things on new PCs as an option. Just include the .EXE file for Folding@Home (or one of the lesser projects :), a link on the desktop and an explanation of what the user can do with his/her idle CPU time. The number crunching power of millions upon millions of PCs wouldn't go to waste.