Democrat judges (who already have rewritten the law, in effect changing the rules of the game after the ball has been put in play)
Did you read the Florida Supreme Court ruling? I think they made an overwhelming case that the early certification "deadline" was not the law.
That said, I think they should have let Katherine Harris pick a new deadline guided by their (correct) reading of the law, rather than picking one themselves.
When I was an undergraduate, I wrote a program that would detect whether a particular move in Minesweeper was safe. It used a recursive search, and couldn't detect safe moves in certain late-game situations where the mine count was relevant, but its play was otherwise perfect.
Since the program could definitively tell whether or not a move was safe, it could detect when a player was *GUESSING*. And so we could hack the program to always reveal a mine in such cases, driving the game weenies insane.:-)
Okay, we never built the search into the game, but we did hack Tetris in a few irritating ways... (As I understand it, Tetris is a lost game anyway: with probability 1, if you play long enough, you will lose, no matter your speed or strategy.)
My worry is that this case will become a precedent cited in the next xyz-really-sucks.com dispute. The fact that the guy made no response may have been the deciding factor, but I wonder if that detail will be overlooked in the future? That is, I wonder if this guy's failure to defend himself cause the screws to tighten on us all.
I agree! Now, why is it that the NASA homepage starts with links to speeches by NASA bureaucrats, and not to these amazing close-up photos of Eros? In fact, there are no links to NEAR anywhere on the homepage. In fact, there are no links to NEAR even on the "Hot Topics" page. Of course, you can still get directly to the bureaucrat speeches from there. Cause that's a pretty hot topic. Probably you'll all want to do that right away. Yeah. I'm thinkin' those speeches are gonna get purdy damn Slashdotted in jusasec here. Uh-huh.
Why is the business / for-fun distinction so damn critical? Justify this statement:
Once you cross the threshold of being a business versus a site that you run for fun you should feel obligated to be as accurate as possible.
Why? *WHY?* What law, what regulation, what principle, what amendment to the constitution, what commandment from god, what lame-ass-line-of-bs makes this true? That's just a total non sequitur.
What's inherently wrong with a business that strives only for so-so accuracy in exchange for other things, such as immediacy, pleasant informality, an opportunity to flame MousePotato, etc? Are you just concerned that it won't be a successful business? Don't fret on that score...
Is is that Rob & Co are now getting paid and its their job? If that's your rationale, then why aren't his obligations a matter between HIM and HIS EMPLOYER? Where do YOU come in? Please tell me what you're getting paid and send me monthly reports so I may decide whether you're working hard enough. I note that you posted at 1:21 PM on a work day.
we deserve to have it be accurate
Why do you *deserve* anything from Slashdot when you get it for free? Please send me $500; I *deserve* it from you.
Criticisms of Slashdot are great and all, but thoughtful criticisms should take into account that (1) it's free (2) Rob works his ass off; you can't ask more of him-- there are only so many hours in the day, and (3) while Slashdot has grown immensely popular, its still produced by Rob and his semi-illiterate gradeschool buddies from Holland, Michigan and that makes it special. Not CNet, not ZDNet, not Salon or NYTimes-- just kooky, quirky, idosyncratic Slashdot. And I, for one, hope that never changes.
I assume you clicked through their agreement. And isn't click-through considered now considered a legally binding form of "digital signature" under that stupid new law? I forget exactly what the agreement said, but I think you already have (inadvertantly) signed something.
Hey! Yo! Over here, guys! We got an expert in JPS!:)
So, uh, could you be persuaded to post some more details?:) Did you, um, look at some kind of Fourier transform (discrete cosine transform?) and look for some tweaking between A and A'? Were you really able to find something??? It seemed to me that they had so many choices-- watermarking individual k-second blocks, say-- that it would be very difficult to reverse engineer their watermarking procedure based on a single example.
My speculation had been that the scheme was cracked by someone with inside knowledge, as there are apparently a lot of folks in SDMI trying to undermine this thing through leaking. Maybe some people did have details of the verification process.
As I understand it, their idea is to have a fragile watermark and a robust watermark in each song on a CD. Ripping to an MP3 will destroy the fragile watermark, but leave the robust watermark intact. A player can refuse to play if it detects this situation. Admitting their unforgability, what role do digital signatures play? Surely they can not be the robust watermark-- one could just clip them. Do digital signatures substitute for the fragile watermark?
(The new Salon article says: "All four technologies in the public test had successful attacks submitted against them." The source is, apparently, only talking about the watermark technologies. I think the Salon author is a little confused on this point.)
Here's a vaguely-related problem that is fun to contemplate over dinner. You've got a square napkin. You can fold it however you want provided that the resulting shape lies in a plane. For example, you could fold a corner over, which reduces the perimeter, and then fold a "sub-corner" back (as students do with homeworks when they don't have a stapler.:-) ), which increases the perimeter again. The question is, can the perimeter of the shape you form by folding ever exceed the perimeter of the original napkin?
The answer turns out to be... Oh! Look at the time! Gotta get going!:-)
I've been looking into watermarking a bit, and I'm less confident about such assertions than I was a few days ago.
In particular, there's this awesome paper online. (Click.pdf or.ps in the upper right corner of the page.) It's remarkable stuff, even if you just look at the pictures. For example, they show a photo before and after watermarking. As you flip back and forth, it's as if the shadows have somehow subtly changed. They do all sorts of crazy stuff like JPEG encode/decode, cutting off parts of the picture, adding noise, photocopying, multiple-watermarking. But none of that destroys the original mark.
Frankly, I'm QUITE surprised that anyone could break watermarks under the conditions of the hacksdmi contest. (Come to think of it, the proposed "technologies" were not all watermarks, right?) My feeling is that if SDMI keeps the watermark verifier in hardware, cracking their scheme could be a real bear.
At least, until some community-friendly engineer anonymously posts details of the verification process on USENET from a public-access terminal.:-)
There seems to be a severe shortage of decent computer people in Boston. I'm in school, but friends in startups are constantly asking if I know any good people that they could hire, since their hiring quotas are going unfilled. One friend talked about getting a half-dozen calls on the day he posted his resume. My academic program has been pillaged by startups; our photo board has a whole section devoted to students prematurely eaten by companies-- as well as several professors.
Peritt had two sentences that I think explain why the Carnivore review is being conducted in such cloak-and-dagger style:
Carnivore is used in sensitive criminal and foreign intelligence investigations.
It is not unreasonable for the Justice Department to assure that the details of confidential criminal investigations or of foreign intelligence methods and procedures will not disclosed to the public.
These are the first overt admissions I've seen that Carnivore is not just a law-enforcement tool. I suspect that the foreign intelligence gathering aspect is what the DoJ, FBI, etc. don't want publicly revealed or even discussed.
For example, perhaps Carnivore does something special with packets that are headed overseas or to foreign embassies. I bet these can be legally tapped at will, much as the NSA is allowed to monitor international (but not domestic) phone calls. I'd guess that scraps of intelligence could frequently be gleaned in this way. Say a Moscow embassy functionary emails his girlfriend back in St. Petersburg and says a tad more than he should to make himself look cool and important. Perhaps Carnivore would gobble this down.
I'm not sure whether NSA conducts industrial espionage as, apparently, some western European intelligence services do. If so, emails from foreign business travelers back home would be a gold mine. This would defintely be hush-hush to a vastly higher degree than banal packet sniffing related to a criminal investigation.
(Of course, why they wouldn't just watch overseas pipes instead of local ISPs isn't clear to me... okay, NO ONE BRING THAT UP, all right? I like my theory.)
Re:Basic error in first algorithm?
on
Does P = NP?
·
· Score: 3
This is not a decision procedure, so there is no "success" or "failure", only termination. The purpose of this algorithm is to construct a sizable bipartite matching, which is subsequently augmented. My point is that such sloppiness in describing a 40-year old procedure does not bode well for the correctness of the paper as a whole. Since posting, I've decrypted the paper through page 10. The entire thing is sloppy, but there it becomes utterly opaque to me.
(I'm a PhD student at MIT specializing in design of combinatorial algorithms and, by coincidence, presently studying fast algorithms for NP-hard problems, e.g. k-SAT, MAX-2-SAT. I referee papers in theoretical CS pretty reguarly. This all to say... yes, I'm familiar with pseudo-code.:-) )
I felt like I understood page 7. This was my read:
Take a graph G. Grab an maximal independent set and call it X0. Delete these vertices from G. Grab another maximal indepdendent set from the (now shrunken) graph G and call it X1. Delete these vertices from G. Continue in this way until G is reduced to the empty graph. This gives a partition of the vertex set of G into parts X0, X1,..., Xm. Then he uses this partition to direct the edges of G, giving him a digraph. In particular, edges always run from lower-index parts to higher-index parts. (So, for example, if (u,v) is an edge of G and u is in part X2 and v is in part X4, then that edge is directed u ---> v.)
That said, I died on page 10, on the "densely stretched" stuff. To the extent that I can decrypt his definitions, his observations look false. For me, the first concrete test of this paper came earlier, though, in his "proof" of Lemma 1. The claim is true, I believe, but the proof seems to be a haphazard set of assertions.
Overall, my feeling is that this paper is almost certainly bogus, and that no one has refuted it only because no one can figure out what the hell he's trying to say. I think Plotnikov needs to find a collaborator that can write coherent English (and mathematics) or else to write the whole thing in his native Russian and submit it to a refereed Russian journal. Posting an unintelligible paper on an obscure web site is just not going to generate proper peer review.
Basic error in first algorithm?
on
Does P = NP?
·
· Score: 4
As far as I can tell, his first algorithm (p. 5-6) contains a simple array-indexing error. Here's my paraphrase:
INPUT: M, an n x n array of booleans.
Let N = n, i = 1.
If N = 0, halt.
If (some condition on the ith row of M)
Let i = i + 1, N = N - 1.
Go to set 2.
else
Let i = i + 1.
Go to step 2.
If the "else" branch is ever taken, then index i will reference a location outside of array M before the procedure halts, right? Here he is just reviewing a 40-year-old algorithm for bipartite matching, but it is disturbing that his very first algorithm contains a basic error.
Taco works his butt off to maintain this site, and people do nothing but ream him. I think I'm a pretty average Slashdot reader, and you know what?
I can decrypt text with incorrect grammar and spelling all by myself. I know this, because a lot of you writing (even insightful ones) comments can't spell worth a damn.
I read at +2 and skip my eyeballs right on past posters that bore me. The moderation system doesn't have to be perfect; I can do some filtering all by myself.
When there's a redundant story, I skip it. When there's an erroneous story, the comments tell me, and I forget it. When there's a spurrious comment in an article, I blow it off.
Why is everything Rob's problem and responsibility? Why is it entirely his job to get everything just all tidy and perfect for pampered little you? Has someone tied you up and forced you to read Slashdot? And, if so, is she good looking?
How about this bit from the IRC log?
[21:47] [CmdrTaco] Some days I just go home so fucking angry because some dickless wonder with no information and a paranoid
fantasty is convinced that I'm the antichrist.,
[21:48] [CmdrTaco] Its great when someone uses a forum that you work so hard to create & maintain to attack you
I wonder what it feels like to work your butt off 10 hours a day and still get dozens of emails a day, every day telling you what a dick you are? I wouldn't take a job like that. And, as far as I know, there's nothing preventing Rob, Jeff, Michael, and co. from saying, "screw this abuse" and dumping Slashdot tomorrow. What do you want from these guys?
Walker (founder of Priceline.com):"Those of us who are creating hundreds and thousands of jobs... are here to tell you that if we slow this engine down because of the study groups you want to have, I'm not sure you're going to be well served in the economy."
New York Times this morning: "Priceline WebHouse Club, which licensed Priceline.com's name-your-own price technology to sell groceries and gasoline, said it would cease operations by early next year. [...] WebHouse's closing deals a blow to Priceline, whose shares fell $2.47, or 26%, to $6.91 in early trading.
You're absolutely right, and I've tried to stick to the facts. If I made any errors, please point them out specifically, and I really will follow up with corrections. I definitely don't want to spread false information. Aside the matter of getting sued, I believe DC sinks on its own demerits. As for being inflamatory-- well, I really believe this is one of those bogus.coms out to make a quick buck from an ill-conceived busines model, and I don't think it's right to let them get away with that. People are going to sink their money into this company-- $100 million if all goes as planned-- and the prevailing opinion around here is that they're going to lose it all. What can I say?
(And rest assured that I used the term "spamming" jokingly above. I've posted at most one comment per IPO discussion forum.)
Hey, I've been doing my part all evening to spam investment boards about the upcoming Digital Convergence IPO. The more the merrier, though! Bust in! Here's a sample:
Digital Convergence (DGTL) recently filed plans for an IPO. This company gives out free barcode scanners (called "CutCat") and accompanying software. The idea is that you can scan things and their software will pull up an appropriate web page in your browser. On the side, they can collect demographic data. For example, they could determine which gender and age group most often scans a certain type of product.
I think this is a horrible company, a must avoid stock for the following reasons:
DGTL gives away CueCat barcode scanners and software, hoping to get money from advertisers and publications. The problem is that their software is inessential: it took folks a matter of hours to write substitute software that reads a barcode without contacting DGTL. So at the key step where they're supposed to cash in, they're completely cut out of the loop! Whoops!
Apparently realizing the enormity of their error, DC has been sending vague, threatening letters to people already distributing alternate software. Unfortunately, these letters appear to be legal bluffs. Decoding software is available on dozens of sites and appears to have no real legal strings attached.
A clearly disconcerted president of the technology group at DGTL fired off a letter showing gross misunderstanding of intellectual property law-- upon which the health of the company critically depends. (Or would depend, were the IP law favorable to their cause-- which it isn't.)
These threatening letters have incensed the open source community-- a group well-qualified to undermine DC's business model by providing alternate software to drive the CueCat, shutting of DC's revenue.
The product raises privacy concerns. You register with DGTL and then every time you scan something, they know it. Apparently DGTL has given assurances about privacy. Then again, they left their entire customer database unguarded for hackers to take. Read their own toned down account. (DGTL has also touted the scanner's "built in encryption", which turned out to consist of XORing each byte with the letter 'C'. I fear these some of the stupidest people ever put on God's Good Earth.)
A key asset that DGTL hopes to develop through the barcode scans is a database of demographic data. There's a problem, however: Digital Convergence has a lot of enemies now. It would be a simple matter for ONE PERSON write a little program that sends fake scans with fake user IDs to DGTLs servers. This could permanently corrupt the demographic database, making it worthless, because-- quite possibly-- there could be no way to distinguish real scans from fakes after the fact.
Just as the company's fundamental business model has fallen under shadow, they file for an IPO. Avoid, avoid, avoid.
These are just my opinions, of course. I did my best to get the facts straight, but I'm not perfect. Additional comments on this corporate disaster slouching toward NASDAQ are available at:
Salon :...there are a million problems with this concept.
Linux World: In the end, the:CueCat is a classic example of a broken business model.
Dr. Dobbs Journal: What ought to scare Digital Convergence more [...] is a database of all CueCat barcodes/URLs, whereby users could go to a specific web page without being tracked.
Dallas Observer:...you can simply drag the scanner 600 or 700 times over bar codes printed next to stories and ads, and presto, you get an error message.
I think Digital Convergence is getting all friendly for the moment because they're about to IPO. Lessee... if we point out all the faults of this company on internet discussion boards and and pull down their initial share price even 10%, that's a $10 million loss for them. I'M ON IT BABY!!!
(Be sure to point out that their demographics database could easily be permanently corrupted by angry hackers feeding them a mass of fake scans. This is an important risk that most investors would overlook.)
what e-signatures will do is make signature fraud substantially more difficult to accomplish
This is a nice post, but like many people here, you're confusing e-signatures (zero security) with digital signatures (cryptographic mechanism). Unfortunately, Congress picked the wrong one to make legally binding as well.:(
At least in my experience, Netmeeting doesn't seem to keep consistency between everyone's whiteboards very well. Drove me nuts, and I eventually gave up.
It's an old observation that important rights cases always involve shady characters. When the rights of sweet, old grandmothers are threatened, it's just not a tough call. The difficult, and therefore important, calls involve characters like Ernesto Miranda, who was arrested for kidnapping and raping a slightly retarded 18-year-old woman.
So to address your question: is this really a case of free speech? Absolutely. The fact that the speakers are associated with low-grade lawbreaking does not in any way detract from the vastly more important free speech issue. Perhaps you've taken your eyes off the ball?
Slashdot Mirror
Democrat judges (who already have rewritten the law, in effect changing the rules of the game after the ball has been put in play)
Did you read the Florida Supreme Court ruling? I think they made an overwhelming case that the early certification "deadline" was not the law.
That said, I think they should have let Katherine Harris pick a new deadline guided by their (correct) reading of the law, rather than picking one themselves.
And since the winner will be a murderer, we can immediately disqualify him for the presidency.
Seems a pretty ideal solution to me.
The date on which projects due "at the end of April" are completed.
When I was an undergraduate, I wrote a program that would detect whether a particular move in Minesweeper was safe. It used a recursive search, and couldn't detect safe moves in certain late-game situations where the mine count was relevant, but its play was otherwise perfect.
Since the program could definitively tell whether or not a move was safe, it could detect when a player was *GUESSING*. And so we could hack the program to always reveal a mine in such cases, driving the game weenies insane. :-)
Okay, we never built the search into the game, but we did hack Tetris in a few irritating ways... (As I understand it, Tetris is a lost game anyway: with probability 1, if you play long enough, you will lose, no matter your speed or strategy.)
My worry is that this case will become a precedent cited in the next xyz-really-sucks.com dispute. The fact that the guy made no response may have been the deciding factor, but I wonder if that detail will be overlooked in the future? That is, I wonder if this guy's failure to defend himself cause the screws to tighten on us all.
I agree! Now, why is it that the NASA homepage starts with links to speeches by NASA bureaucrats, and not to these amazing close-up photos of Eros? In fact, there are no links to NEAR anywhere on the homepage. In fact, there are no links to NEAR even on the "Hot Topics" page. Of course, you can still get directly to the bureaucrat speeches from there. Cause that's a pretty hot topic. Probably you'll all want to do that right away. Yeah. I'm thinkin' those speeches are gonna get purdy damn Slashdotted in jusasec here. Uh-huh.
Why is the business / for-fun distinction so damn critical? Justify this statement:
Once you cross the threshold of being a business versus a site that you run for fun you should feel obligated to be as accurate as possible.
Why? *WHY?* What law, what regulation, what principle, what amendment to the constitution, what commandment from god, what lame-ass-line-of-bs makes this true? That's just a total non sequitur.
What's inherently wrong with a business that strives only for so-so accuracy in exchange for other things, such as immediacy, pleasant informality, an opportunity to flame MousePotato, etc? Are you just concerned that it won't be a successful business? Don't fret on that score...
Is is that Rob & Co are now getting paid and its their job? If that's your rationale, then why aren't his obligations a matter between HIM and HIS EMPLOYER? Where do YOU come in? Please tell me what you're getting paid and send me monthly reports so I may decide whether you're working hard enough. I note that you posted at 1:21 PM on a work day.
we deserve to have it be accurate
Why do you *deserve* anything from Slashdot when you get it for free? Please send me $500; I *deserve* it from you.
Criticisms of Slashdot are great and all, but thoughtful criticisms should take into account that (1) it's free (2) Rob works his ass off; you can't ask more of him-- there are only so many hours in the day, and (3) while Slashdot has grown immensely popular, its still produced by Rob and his semi-illiterate gradeschool buddies from Holland, Michigan and that makes it special. Not CNet, not ZDNet, not Salon or NYTimes-- just kooky, quirky, idosyncratic Slashdot. And I, for one, hope that never changes.
I assume you clicked through their agreement. And isn't click-through considered now considered a legally binding form of "digital signature" under that stupid new law? I forget exactly what the agreement said, but I think you already have (inadvertantly) signed something.
Hey! Yo! Over here, guys! We got an expert in JPS! :)
So, uh, could you be persuaded to post some more details? :) Did you, um, look at some kind of Fourier transform (discrete cosine transform?) and look for some tweaking between A and A'? Were you really able to find something??? It seemed to me that they had so many choices-- watermarking individual k-second blocks, say-- that it would be very difficult to reverse engineer their watermarking procedure based on a single example.
My speculation had been that the scheme was cracked by someone with inside knowledge, as there are apparently a lot of folks in SDMI trying to undermine this thing through leaking. Maybe some people did have details of the verification process.
As I understand it, their idea is to have a fragile watermark and a robust watermark in each song on a CD. Ripping to an MP3 will destroy the fragile watermark, but leave the robust watermark intact. A player can refuse to play if it detects this situation. Admitting their unforgability, what role do digital signatures play? Surely they can not be the robust watermark-- one could just clip them. Do digital signatures substitute for the fragile watermark?
(The new Salon article says: "All four technologies in the public test had successful attacks submitted against them." The source is, apparently, only talking about the watermark technologies. I think the Salon author is a little confused on this point.)
Here's a vaguely-related problem that is fun to contemplate over dinner. You've got a square napkin. You can fold it however you want provided that the resulting shape lies in a plane. For example, you could fold a corner over, which reduces the perimeter, and then fold a "sub-corner" back (as students do with homeworks when they don't have a stapler. :-) ), which increases the perimeter again. The question is, can the perimeter of the shape you form by folding ever exceed the perimeter of the original napkin?
The answer turns out to be... Oh! Look at the time! Gotta get going! :-)
I've been looking into watermarking a bit, and I'm less confident about such assertions than I was a few days ago.
In particular, there's this awesome paper online. (Click .pdf or .ps in the upper right corner of the page.) It's remarkable stuff, even if you just look at the pictures. For example, they show a photo before and after watermarking. As you flip back and forth, it's as if the shadows have somehow subtly changed. They do all sorts of crazy stuff like JPEG encode/decode, cutting off parts of the picture, adding noise, photocopying, multiple-watermarking. But none of that destroys the original mark.
Frankly, I'm QUITE surprised that anyone could break watermarks under the conditions of the hacksdmi contest. (Come to think of it, the proposed "technologies" were not all watermarks, right?) My feeling is that if SDMI keeps the watermark verifier in hardware, cracking their scheme could be a real bear.
At least, until some community-friendly engineer anonymously posts details of the verification process on USENET from a public-access terminal. :-)
There seems to be a severe shortage of decent computer people in Boston. I'm in school, but friends in startups are constantly asking if I know any good people that they could hire, since their hiring quotas are going unfilled. One friend talked about getting a half-dozen calls on the day he posted his resume. My academic program has been pillaged by startups; our photo board has a whole section devoted to students prematurely eaten by companies-- as well as several professors.
Peritt had two sentences that I think explain why the Carnivore review is being conducted in such cloak-and-dagger style:
These are the first overt admissions I've seen that Carnivore is not just a law-enforcement tool. I suspect that the foreign intelligence gathering aspect is what the DoJ, FBI, etc. don't want publicly revealed or even discussed.
For example, perhaps Carnivore does something special with packets that are headed overseas or to foreign embassies. I bet these can be legally tapped at will, much as the NSA is allowed to monitor international (but not domestic) phone calls. I'd guess that scraps of intelligence could frequently be gleaned in this way. Say a Moscow embassy functionary emails his girlfriend back in St. Petersburg and says a tad more than he should to make himself look cool and important. Perhaps Carnivore would gobble this down.
I'm not sure whether NSA conducts industrial espionage as, apparently, some western European intelligence services do. If so, emails from foreign business travelers back home would be a gold mine. This would defintely be hush-hush to a vastly higher degree than banal packet sniffing related to a criminal investigation.
(Of course, why they wouldn't just watch overseas pipes instead of local ISPs isn't clear to me... okay, NO ONE BRING THAT UP, all right? I like my theory.)
This is not a decision procedure, so there is no "success" or "failure", only termination. The purpose of this algorithm is to construct a sizable bipartite matching, which is subsequently augmented. My point is that such sloppiness in describing a 40-year old procedure does not bode well for the correctness of the paper as a whole. Since posting, I've decrypted the paper through page 10. The entire thing is sloppy, but there it becomes utterly opaque to me.
(I'm a PhD student at MIT specializing in design of combinatorial algorithms and, by coincidence, presently studying fast algorithms for NP-hard problems, e.g. k-SAT, MAX-2-SAT. I referee papers in theoretical CS pretty reguarly. This all to say... yes, I'm familiar with pseudo-code. :-) )
I felt like I understood page 7. This was my read:
Take a graph G. Grab an maximal independent set and call it X0. Delete these vertices from G. Grab another maximal indepdendent set from the (now shrunken) graph G and call it X1. Delete these vertices from G. Continue in this way until G is reduced to the empty graph. This gives a partition of the vertex set of G into parts X0, X1, ..., Xm. Then he uses this partition to direct the edges of G, giving him a digraph. In particular, edges always run from lower-index parts to higher-index parts. (So, for example, if (u,v) is an edge of G and u is in part X2 and v is in part X4, then that edge is directed u ---> v.)
That said, I died on page 10, on the "densely stretched" stuff. To the extent that I can decrypt his definitions, his observations look false. For me, the first concrete test of this paper came earlier, though, in his "proof" of Lemma 1. The claim is true, I believe, but the proof seems to be a haphazard set of assertions.
Overall, my feeling is that this paper is almost certainly bogus, and that no one has refuted it only because no one can figure out what the hell he's trying to say . I think Plotnikov needs to find a collaborator that can write coherent English (and mathematics) or else to write the whole thing in his native Russian and submit it to a refereed Russian journal. Posting an unintelligible paper on an obscure web site is just not going to generate proper peer review.
As far as I can tell, his first algorithm (p. 5-6) contains a simple array-indexing error. Here's my paraphrase:
INPUT: M, an n x n array of booleans.
If the "else" branch is ever taken, then index i will reference a location outside of array M before the procedure halts, right? Here he is just reviewing a 40-year-old algorithm for bipartite matching, but it is disturbing that his very first algorithm contains a basic error .
Or am I missing something? :-)
Taco works his butt off to maintain this site, and people do nothing but ream him. I think I'm a pretty average Slashdot reader, and you know what?
Why is everything Rob's problem and responsibility? Why is it entirely his job to get everything just all tidy and perfect for pampered little you? Has someone tied you up and forced you to read Slashdot? And, if so, is she good looking?
How about this bit from the IRC log?
[21:47] [CmdrTaco] Some days I just go home so fucking angry because some dickless wonder with no information and a paranoid fantasty is convinced that I'm the antichrist.,
[21:48] [CmdrTaco] Its great when someone uses a forum that you work so hard to create & maintain to attack you
I wonder what it feels like to work your butt off 10 hours a day and still get dozens of emails a day, every day telling you what a dick you are? I wouldn't take a job like that. And, as far as I know, there's nothing preventing Rob, Jeff, Michael, and co. from saying, "screw this abuse" and dumping Slashdot tomorrow. What do you want from these guys?
Walker (founder of Priceline.com): "Those of us who are creating hundreds and thousands of jobs ... are here to tell you that if we slow this engine down because of the study groups you want to have, I'm not sure you're going to be well served in the economy."
New York Times this morning: "Priceline WebHouse Club, which licensed Priceline.com's name-your-own price technology to sell groceries and gasoline, said it would cease operations by early next year. [...] WebHouse's closing deals a blow to Priceline, whose shares fell $2.47, or 26%, to $6.91 in early trading.
Thanks for the constructive feedback. I'll adjust my postings with your suggestions in mind.
If you would stick to the facts, that would help.
You're absolutely right, and I've tried to stick to the facts. If I made any errors, please point them out specifically, and I really will follow up with corrections. I definitely don't want to spread false information. Aside the matter of getting sued, I believe DC sinks on its own demerits. As for being inflamatory-- well, I really believe this is one of those bogus .coms out to make a quick buck from an ill-conceived busines model, and I don't think it's right to let them get away with that. People are going to sink their money into this company-- $100 million if all goes as planned-- and the prevailing opinion around here is that they're going to lose it all. What can I say?
(And rest assured that I used the term "spamming" jokingly above. I've posted at most one comment per IPO discussion forum.)
Hey, I've been doing my part all evening to spam investment boards about the upcoming Digital Convergence IPO. The more the merrier, though! Bust in! Here's a sample:
Digital Convergence (DGTL) recently filed plans for an IPO. This company gives out free barcode scanners (called "CutCat") and accompanying software. The idea is that you can scan things and their software will pull up an appropriate web page in your browser. On the side, they can collect demographic data. For example, they could determine which gender and age group most often scans a certain type of product.
I think this is a horrible company, a must avoid stock for the following reasons:
These are just my opinions, of course. I did my best to get the facts straight, but I'm not perfect. Additional comments on this corporate disaster slouching toward NASDAQ are available at:
I think Digital Convergence is getting all friendly for the moment because they're about to IPO. Lessee... if we point out all the faults of this company on internet discussion boards and and pull down their initial share price even 10%, that's a $10 million loss for them. I'M ON IT BABY!!!
(Be sure to point out that their demographics database could easily be permanently corrupted by angry hackers feeding them a mass of fake scans. This is an important risk that most investors would overlook.)
what e-signatures will do is make signature fraud substantially more difficult to accomplish
This is a nice post, but like many people here, you're confusing e-signatures (zero security) with digital signatures (cryptographic mechanism). Unfortunately, Congress picked the wrong one to make legally binding as well. :(
At least in my experience, Netmeeting doesn't seem to keep consistency between everyone's whiteboards very well. Drove me nuts, and I eventually gave up.
It's an old observation that important rights cases always involve shady characters. When the rights of sweet, old grandmothers are threatened, it's just not a tough call. The difficult, and therefore important, calls involve characters like Ernesto Miranda, who was arrested for kidnapping and raping a slightly retarded 18-year-old woman.
So to address your question: is this really a case of free speech? Absolutely. The fact that the speakers are associated with low-grade lawbreaking does not in any way detract from the vastly more important free speech issue. Perhaps you've taken your eyes off the ball?