Prior to reading the transcript above in Montfort's book, I'd heard of SHRDLU only in passing. It deserves a lot more attention than it's received.
SHRDLU received plenty of attention at the time. Modern AI is pretty much based on Winograd's work. SHRDLU was the first of the expert systems, programs with relatively simple pattern/action reasoning systems mated to large databases of expert domain knowledge. They were the basis of the great commercial AI scare of the 1980s, into which many zillions of venture capital was poured, and from which sprang, well, not much.
It's not Kerry going after them, it's the photographers (and their agents) whose work was misappropriated. In any case, Kerry wouldn't have standing to object on copyright or DMCA grounds. He might have a libel case, though.
Pixar has a double handful of these, for Renderman, for digital film printers, and several others. Two of them (for designing Pixar's animation system and for the first digital compositing system) are on the mantle at my house.
OSDL is a non-profit organization, so donations should be tax-deductible. (I can't actually tell from their web site if they're 510c3 or just an Oregon nonprofit, but I bet they'd do the federal paperwork if they saw a $250K donation coming.)
He might mean "A History of Science and Engineering in the Bell System", a series of many volumes. A google search reveals this link, which gives these isbns: 0-932764-07-X, 0-932764-06-1. I'm pretty sure there were 5 or 6 volumes in the series.
In a free market they will be motivated to sell cheaper than their neighbors.
No, they'll be motivated to maximize their receipts. In a free market for commodities, the usual way to do this is to collude with other suppliers. In the US, by contrast, the Sherman act encourages selling cheaper than competitors. Your idea would work if not for a jursidiction problem.
Knuth retired from his position at Stanford as part of a program of eliminating all commitments that might keep him from working on the remaining volumes of The Art of Computer Programming. He certainly hasn't stopped work. In fact, he's already circulating drafts of several sections of Volume 4, his first new work (other than revision of volumes 1-3) on TAOCP in many years. Progress report here.
In C, functions return rvalues but not lvalues. (In something a little more like English: the value returned by a function isn't in a place you can take the address of.) But ++ requires its operand to be an lvalue.
Right. So the fix is not to do that -- i.e. don't use simple transponders, but populate the satellite with spread spectrum relays that cannot be jammed by a strong unmodulated carrier because they only watch that frequency for a few microseconds at a time, and can use forward error correction to correct for the interference. In fact, no signal of plausible amplitude that is unaware of the spreading sequence (which possibility we circumvent with a little cryptography) can successfully jam the transmission -- the jammer has to transmit on a substantial fraction of the frequencies that the spread spectrum signal uses, with higher power on each of them than the legit signal uses. Say we use a frequency-hopping system with 10 thousand frequencies (not hard in the microwave bands), at a power of 100 watts. The jammer has to jam a good fraction of the frequencies to have any success, so he has to build a transmitter whose power output is a good fraction of a megawatt, a proposition that is out of the range of most prospective jammers. And if it's not, we just broaden the spectrum of our frequency-hopper by a few factors of 10 until we exceed the bad guy's budget.
The satellite in question went into service in 1999, thirteen years after Capt Midnight demonstrated how to take over an uplink. It's not like they haven't had time to figure this out, or that the expense involved is large compared to the cost of a communication satellite. So why haven't they done it? The only plausible explanations are inertia and incompetence.
As has been pointed out, people have been jamming satellite uplinks since Capt Midnite & HBO. So why haven't the satellite folks gone to jam-resistant technology? (I know, it's because of the essential laziness of corporate culture -- geez, it was a rhetorical question.) Spread spectrum is essentially unjammable, if done right (i.e. with cryptographically generated spreading sequences or some such cryptogeek mumbojumbo.)
B didn't come from anything called A, but from a language called bon, named for Ken Thompson's wife Bonnie. (At least, that's what Ken says, but he's famous for pulling the legs of people who drool over stupid trivia from his past.) The inspiration for bon was Martin Richards' BCPL, a stripped version of Christopher Strachey's [et al] CPL (Combined Programming Language, I think; the B in BCPL is for Bootstrap or Basic, sources differ.) It doesn't stretch the truth too far to think of B as an even-more stripped down BCPL.
I've been told (by a Babylonian scholar) that there are more people alive today who read and write Babylonian cuneiform than at any other time in history.
When it was a living language, the world's population was tiny and the literacy rate microscopic. The literate of ancient Babylon are far outnumbered by the linguistics undergraduates who study cuneiform today.
Not Lucky, Forchini
on
Reflections
·
· Score: 2, Informative
Bob Lucky, who doesn't even work at Bell Labs any more (he bounced over to Telcordia), didn't do this work, Gerard Foschini did, as you could tell by reading the article.
Ideas and information are not the only intellectual activities, and we all know that copyright doesn't protect them, except incidentally. Expression of ideas and information, which copyright does protect, is also an intellectual activity.
30000 is roughly the square root of 10^9 (the factor that 90 dB refers to.) The square root is because signal power varies as the square of signal amplitude.
.01/667 was a typo for.1/667. If all the people taking me to task for not checking my calculations had checked for themselves, they would have seen that.
I used the 90db figure because audiophiles still claim that vinyl sounds better than CDs, for which the available dynamic range is rougly 96 dB. If I'd used 70 dB (the figure bandied about here) instead, the feature size would be 50 nm and the conclusion would be no different.
A vinyl LP is 12 inches in diameter, and has a label area in the middle that's about 4 inches in diameter. So the area containing the spiral groove is about 4 inches wide. That's about 10 centimeters. An LP side typically has a little more than 20 minutes of music on it. It rotates at 33 1/3 RPM, so the groove spirals around roughly 667 times. So the width of the groove is roughly.01/667 meters, which is 150 microns. The signal (on a monaural record, stereo is more complicated!) is recorded by wiggling the groove from side to side in that 150 micron space. To reproduce a signal whose dynamic range is 90 dB, the smallest excursions have to be roughly 1/30000 of the maximum amplitude. 150/30000 microns is 5 nanometers.
Think your scanner has that much resolution? Guess again -- 1200 dpi is roughly 21 microns, off by a factor of 100.
Note that 5 nanometers is way smaller than the wavelength of visible light (roughly 750 to 350 nm), so those laser turntables everyone is talking about don't work very well either, unless they've got x-ray lasers in them.
Slashdot Mirror
But as for the millions, who actually get's the money?
The apostrophe fairy?
Lucida is a Bigelow & Holmes design. Was Bitstream their client?
SHRDLU received plenty of attention at the time. Modern AI is pretty much based on Winograd's work. SHRDLU was the first of the expert systems, programs with relatively simple pattern/action reasoning systems mated to large databases of expert domain knowledge. They were the basis of the great commercial AI scare of the 1980s, into which many zillions of venture capital was poured, and from which sprang, well, not much.
Probably Wired asked a dozen designers to improve google, eight said "What? It's perfect already!" and the other four made it into the article.
It's not Kerry going after them, it's the photographers (and their agents) whose work was misappropriated. In any case, Kerry wouldn't have standing to object on copyright or DMCA grounds. He might have a libel case, though.
Pixar has a double handful of these, for Renderman, for digital film printers, and several others. Two of them (for designing Pixar's animation system and for the first digital compositing system) are on the mantle at my house.
OSDL is a non-profit organization, so donations should be tax-deductible. (I can't actually tell from their web site if they're 510c3 or just an Oregon nonprofit, but I bet they'd do the federal paperwork if they saw a $250K donation coming.)
He might mean "A History of Science and Engineering in the Bell System", a series of many volumes. A google search reveals this link, which gives these isbns: 0-932764-07-X, 0-932764-06-1. I'm pretty sure there were 5 or 6 volumes in the series.
No, they'll be motivated to maximize their receipts. In a free market for commodities, the usual way to do this is to collude with other suppliers. In the US, by contrast, the Sherman act encourages selling cheaper than competitors. Your idea would work if not for a jursidiction problem.
100MHz was the absolute limit for the speed of a CPU.
This isn't true? Does that mean I have to upgrade my 90 MHz Pentiums now?
Knuth retired from his position at Stanford as part of a program of eliminating all commitments that might keep him from working on the remaining volumes of The Art of Computer Programming. He certainly hasn't stopped work. In fact, he's already circulating drafts of several sections of Volume 4, his first new work (other than revision of volumes 1-3) on TAOCP in many years. Progress report here.
I forgot to say: a way for mapsize(mp)++; to be legal C is for mapsize to be a #define.
In C, functions return rvalues but not lvalues. (In something a little more like English: the value returned by a function isn't in a place you can take the address of.) But ++ requires its operand to be an lvalue.
Right. So the fix is not to do that -- i.e. don't use simple transponders, but populate the satellite with spread spectrum relays that cannot be jammed by a strong unmodulated carrier because they only watch that frequency for a few microseconds at a time, and can use forward error correction to correct for the interference. In fact, no signal of plausible amplitude that is unaware of the spreading sequence (which possibility we circumvent with a little cryptography) can successfully jam the transmission -- the jammer has to transmit on a substantial fraction of the frequencies that the spread spectrum signal uses, with higher power on each of them than the legit signal uses. Say we use a frequency-hopping system with 10 thousand frequencies (not hard in the microwave bands), at a power of 100 watts. The jammer has to jam a good fraction of the frequencies to have any success, so he has to build a transmitter whose power output is a good fraction of a megawatt, a proposition that is out of the range of most prospective jammers. And if it's not, we just broaden the spectrum of our frequency-hopper by a few factors of 10 until we exceed the bad guy's budget.
The satellite in question went into service in 1999, thirteen years after Capt Midnight demonstrated how to take over an uplink. It's not like they haven't had time to figure this out, or that the expense involved is large compared to the cost of a communication satellite. So why haven't they done it? The only plausible explanations are inertia and incompetence.
As has been pointed out, people have been jamming satellite uplinks since Capt Midnite & HBO. So why haven't the satellite folks gone to jam-resistant technology? (I know, it's because of the essential laziness of corporate culture -- geez, it was a rhetorical question.) Spread spectrum is essentially unjammable, if done right (i.e. with cryptographically generated spreading sequences or some such cryptogeek mumbojumbo.)
B didn't come from anything called A, but from a language called bon, named for Ken Thompson's wife Bonnie. (At least, that's what Ken says, but he's famous for pulling the legs of people who drool over stupid trivia from his past.) The inspiration for bon was Martin Richards' BCPL, a stripped version of Christopher Strachey's [et al] CPL (Combined Programming Language, I think; the B in BCPL is for Bootstrap or Basic, sources differ.) It doesn't stretch the truth too far to think of B as an even-more stripped down BCPL.
I've been told (by a Babylonian scholar) that there are more people alive today who read and write Babylonian cuneiform than at any other time in history.
When it was a living language, the world's population was tiny and the literacy rate microscopic. The literate of ancient Babylon are far outnumbered by the linguistics undergraduates who study cuneiform today.
This isn't exactly a new view. James Gosling's classic Phase Relationships in the Standardization Process is already 13 years old.
Bob Lucky, who doesn't even work at Bell Labs any more (he bounced over to Telcordia), didn't do this work, Gerard Foschini did, as you could tell by reading the article.
Ideas and information are not the only intellectual activities, and we all know that copyright doesn't protect them, except incidentally. Expression of ideas and information, which copyright does protect, is also an intellectual activity.
There's a pair red pins (unsecured WAPs) on the west side of University Avenue between Dundas and Queen, right about where the US Consulate is.
30000 is roughly the square root of 10^9 (the factor that 90 dB refers to.) The square root is because signal power varies as the square of signal amplitude.
.01/667 was a typo for .1/667. If all the people taking me to task for not checking my calculations had checked for themselves, they would have seen that.
I used the 90db figure because audiophiles still claim that vinyl sounds better than CDs, for which the available dynamic range is rougly 96 dB. If I'd used 70 dB (the figure bandied about here) instead, the feature size would be 50 nm and the conclusion would be no different.
Uh. It's much more than a factor of 100 off. It's more like a factor of 4200. I divided the wrong two numbers when I wrote that.
A vinyl LP is 12 inches in diameter, and has a label area in the middle that's about 4 inches in diameter. So the area containing the spiral groove is about 4 inches wide. That's about 10 centimeters. An LP side typically has a little more than 20 minutes of music on it. It rotates at 33 1/3 RPM, so the groove spirals around roughly 667 times. So the width of the groove is roughly .01/667 meters, which is 150 microns. The signal (on a monaural record, stereo is more complicated!) is recorded by wiggling the groove from side to side in that 150 micron space. To reproduce a signal whose dynamic range is 90 dB, the smallest excursions have to be roughly 1/30000 of the maximum amplitude. 150/30000 microns is 5 nanometers.
Think your scanner has that much resolution? Guess again -- 1200 dpi is roughly 21 microns, off by a factor of 100.
Note that 5 nanometers is way smaller than the wavelength of visible light (roughly 750 to 350 nm), so those laser turntables everyone is talking about don't work very well either, unless they've got x-ray lasers in them.