The Osprey is a spectacular concept, and the XV-15 was a completely successful prototype. It appears to me that the jump made to the production Osprey added so much complexity, and cut the tolerances so fine, that the plane is in serious trouble. Adding the folding blades, the pivoting wing, the high-pressure hydraulic system, and so forth -- it was more that could be done in one go, and the program is suffering for it.
Tilt-rotors will revolutionize aviation, there's no doubt about it. The record of the Osprey, though, really is distinctly worse than other programs that get to this stage. It's not just PR. Falsifying maintenance records didn't help, though.
I think that the Lockheed engineers wanted to build a bigger, heavier plane, and had to go to the lift fan. Direct lift on a plane as heavy as the X-35 just wasn't going to work. So, they had to try something radical. Well, they convinced some people. If it ends up working, well, great -- I'll eat my hat. Won't be the first time.
I thought that the Boeing plane should have won the competition, mostly because it fulfilled the specification better; while being smaller, lighter, and immeasurably simpler. The Boeing plane didn't take off vertically, it's true, but that is also not in the specification -- it's not what was asked for. Similarly, there was no line-item for aesthetics. The Boeing direct-lift concept is the same that powers the Harrier, and is the only demonstrated successful direct lift formula. The clutch-driven lift fan is an Osprey-scale debacle waiting to happen -- mechanically clutching in 40,000 HP in a few seconds with an extremely lightweight gearbox is, I believe, untenable. They finally got it to work for a few tests, but there were a number of fairly spectacular failures along the way. The Boeing design lets the pilot shift from forward to vertical thrust and back again in a few seconds, at will, and they did it more than 100 times during the flight test program -- the Lockheed one was only clutched a handful of times.
The very wide-chord wing of the Boeing design is good for a number of structural, aerodynamic, and stealth reasons. Unfortunately, Boeing elected to change the design for the actual plane to a separate tail, rather than the delta wing -- Lockheed partisans claimed (rightly, IMHO) that this meant that the demonstrator that Boeing flew wasn't really representative of the final plane.
The one terrific thing that the Lockheed design has, the one true aerodynamic innovation, is the bump intake. There's a big bump right in front of the intakes on the Lockheed plane; it performs all of the functions of the typical intake splitter plate, purging the boundary layer, with a far more elegant, lighter, simpler, stealthier, easier-to-maintain design. Hats off to the engineers that came up with that.
I think that the Boeing design is prettier, too, but that's just me -- I'm a low-aspect ratio kind of guy.
If you manage to get through the slashdotting, the story in the tecchannel web pages is amazing. The prototype Clawhammer, while limited to 800 MHz, performed shockingly well on the few, but varied, benchmarks they subjected it to. It's interesting that both Intel and AMD teach the same lesson, that MHz doesn't determine performance. Unfortunately for Intel, they demonstrate it by the P4 not running as fast as the MHz would imply, where the AMD chips run far faster than MHz would imply.
I know that it wasn't you people that did it, but I have to give IBM major props for that ad. For those of you that haven't seen it, there are two guys in suits watching the IBM team play basketball. There's "Middleware", "Applications", and, of course, "Linux", a tall, lumbering, bear of a basketball player.
One guy says to the other, "You know, that guy Linux doesn't get paid a cent for this"
and the other guy says "No! Why does he do it then?"
From the first hundred pages or so, it doesn't seem terribly revolutionary. I read the stunning Cellular Automata Machines by Toffoli and Margolus back when it was first issued in the late eighties, and it seemed far more exciting, and far less egomaniacal.
But, I'll continue to plow on, and see what develops. From the great review above, it does appear that it gets better soon.
Back in 1988, Graham Walters and I, at PDI, built a system in collaboration with Jim Henson and his creature shop, and with Kirk Thatcher (who since went on to greater things at Henson), to build the 'Waldo' character for The Jim Henson Hour. The idea was that this character would be controlled by a waldo, which would sense the position, orientation, and mouth angles, and display the character, blue-screen keyed onto the image on the screen in real time.
The beauty of this system is that Henson puppeteers always work by watching their images on TV monitors, so this kind of digital character wasn't even second-nature to them -- it was exactly how they'd been performing characters all along. Among the nice things about Waldo is that he didn't have to hang out at the bottom of the screen with all of the rest of the other puppets. This system was implemented and run on an old Power Series SGI borrowed from Sheridan College.
I saw Davey Goelz (Gonzo, and others) at Siggraph this year, at the Henson booth on the show floor. They're selling a somewhat improved version of the same waldo mechanism that we used 14 years ago. Davey got us out of a jam on the first Henson Hour show, as somehow we lost the mouth-opening information from the tracks that Henson recorded on the set. Davey lived right near us at PDI, and came down and laid those back in, mimicing Jim's style. I don't think that he ever found our, and it's tragically too late now.
Anyway, The Jim Henson Hour was too good, and perhaps a little too different, for American TV, and only 12 episodes were ever made.
Pulson (and its predecessor company, Time Domain) has been desparately trying to commercialize this technology for radio communication for years. More than five years ago they demonstrated a few-milliwatt UWB radio with 100-mile range. They have mostly been held back by patents taken out by Lawrence Livermore. Livermore claims to have invented all of this stuff, and has been rediculously rough on licensing. Also, the FCC has been unclear until very recently on how it would license UWB.
Aetherwire has attempting to use UWB technology to build localizers, basically extremely short range, extremely low-power peer-to-peer short-range version of GPS. The localizers would all cooperate at keeping track of where the other ones were within a few hundred meter radius. If you've read A Deepness in the Sky by Vernor Vinge, it's all about localizers.
Now that the FCC has cleared the way, I expect to see tremendous progress in UWB. It's going to revolutionize many fields, from radio to positioning to radar.
I know that I'm adding this comment far to late to get noticed, but it must be said. I'm certain that the only reason that MS is making this announcment, and certainly the reason that they are making it now is to impress Colleen Kollar-Kotelly -- saying that they are not trying to take over the world, after all.
thad
Europe's Galileo system is a great example
on
Space Wars
·
· Score: 2
One of the most important space assets is a global positioning system. So far, the US has allowed free use of our DoD-sponsored GPS, but everybody knows that policy could change at some time in the future. The Europeans felt that without their own system, their soverignty was threatened, so they proposed their own system, called Galileo [must have been some Europoean guy.]
Anyway, the US is all in a sweat about this, and has put quite a bit if pressure on the Europeans to get them to forego this system. The US says that it's not needed, that it would interfere with our system, that it would destabilize the planet -- basically the entire bag of boogeymen. To it's credit, Europe has recently reversed course and decided that they would create this system on their own. Germany had been leaning toward the US position, but just changed their mind.
Galileo is not very well defined yet (even basic things like how many satellites, and which orbit configuration), but most people expect that it will be somewhat better than the current GPS system -- although the US insists that it's super-duper GPSII system will be better than Galileo, whatever Galileo ends up being.
LCD's really are dramatically better than they used to be. Mike Massee's statement above was certainly true a few years ago, but it is no longer the case. At Illusion Arts, many of the matte painters are using LCD screens. Some of these people went straight from real paint to LCDs without ever seeing CRTs.
The last time I bought an ATI card I was shocked by the fact that the Linux driver had incredibly bad 2D performance. ATI had paid the Precision Insight folks to write the 3D driver, but didn't include any money for accelerated 2D performance.
I don't know if this has been remedied since then, this was the status 6 months ago.
We live about 100 yards from the Mulholland Drive mentioned at the end of the article, right in the middle of Los Angeles. Interestingly, and frustratingly, even though this part of Bel Air and Beverly Hills is full of people who would desparately like to have broadband access, there is none. No DSL, no cable modems. ATT has pulled out its fixed-wireless system. Metricom of course went belly-up.
Adelphia would be our cable modem provider. They've been busily laying cable for the last year, and have all but completed their network. Now I read in this story, Adelpha claims that it being crippled by DWP, because they can't get power to their network.
I wonder if the Department of Water and Power sees Adelphia as competition, and is inhibiting them in the obvious way. Or, this might be another case where you shouldn't attribute to malice what can equally be explained as bumbling by a cable company.
It will be interesting. Adelphia claims that they'll light up the fibers here within the next month or so. I can't wait.
thad
Showtime was our big FX film for the year
on
Review: Showtime
·
· Score: 5, Interesting
This post will try to make Showtime qualify for the 'News for Nerds' motto of Slashdot.
You might not think of this as a big FX movie, after all, it's just a cop buddy movie, right? Well, it turns out that my little company did over 140 digital FX shots for this film. This demonstrates among other things that computer graphics effects are a big part of almost any film these days -- it gives the director and writers freedom to shoot things more effectively, and allows the director and writer the freedom to improve the film in the post-production phase.
I'll try not to reveal too many plot points to those of you who haven't seen the film -- but if you don't want to know anything about the film you can stop reading right here.
Ok. The biggest thing that we did was build the environment outside the penthouse -- buildings, reflections, and helicopters. The challenge was to keep that environment alive (without the cliche flocks of pigeons.) We did this mostly by observing that windows in skyscrapers flex somewhat in the wind, so that the distortions of the reflections in the windows are always changing. There are a few places that you can see near the ground, and we added little sparkles of light from car windshields, things like that.
We also added a bunch of lasers to the guns, and a bunch of sparks in the gun show sequence. Finally, we did all the 'videoization' and changed the license plate of Eddie Murphy's car for some shots. There were a few dozen other little things here and there, but that's the majority of the shots. By the way, in reality there is a heliport on the roof of the Westin Bonaventure in downtown Los Angeles.
All of this work was done by a team of 10 or so people over about six months. We used a combination of SGI and Linux boxes to do the animation design, and our render garden (it's too small to be a proper render farm) of Linux boxes to do all of the batch rendering.
I've told my employees that if I ever miss a paycheck, they should quit immediately, regardless of what I say at the time. I tell them not to believe me if I say it will get better. I've never seen things get better for a company once they stop being able to make payroll. I don't suppose that the visual effects business is any different than any other business in that respect. Once you start digging a hole it becomes increasingly impossible to ever get out of it.
The problem is, that once a company starts foundering, the founders often begin to lose touch with reality and start making promises that they can't keep -- whether or not they know that is not really an issue. The hole is not only financial (although that's a big enough hurdle on its own) but it's also bad will, that is, the accumulated acrimony festering within the company.
I have been gently pressing the organizers of the 'world conference' about getting the original copy of Mundie's speech, but so far I have been unable to get it. They claimed, at first, to be transcribing the speech, and that it would be available at the media part of the site, but so far it hasn't appeared.
I really want to see the original source, as I believe that it's quite likely that Mundie's reported words are not particularly accurate, and they are surely quoted out of contest. I'm most interested, of course, because I think that the original text is, if anything, more strident and open for redicule.
I've got an email log of my conversations with the World Conference orgranzers that I'd be willing to share with anybody, on request, just send an email. Perhaps with a few more people asking we can get the transcript.
I love cryptome, I read it every day. That said, the report John Young has put up is the unclassified version; although he's trolling for the classified version.
Hey Taco, any idea when the results of the John Young questions will be posted? It's been six months or so...
thad
My favorite space hack
on
Hack in Space
·
· Score: 4, Informative
About five years ago, Hughes launched a communication satellite. To be useful, these have to reach a geosynchronous orbit. It turns out that the most efficient way to get to from a low-earth orbit to a geosynchronous orbit is to fire a rocket twice [or so it was thought.] The first firing raises the apogee of the orbit to the geosynchronous altitude; so that the satellite is in a very elliptical orbit. Then, when the satellite is at apogee, you fire the rocket again to circularize the orbit. Usually this same motor changes the plane of the orbit as well. Most satellites are launched into orbits inclined to the equator somewhat, and geosynchronous satellites have to be over the equator. It is most efficient to make the orbital plane change at apogee, too.
There was a rash of apogee kick motor failures, and in this particular satellite the motor failed, leaving the satellite in a uselessly ellipitical orbit. There were small thrusters on the satellite which were to be used for station-keeping (small orbital adjustments) but it didn't have nearly enough propellant to raise the perigee. Hughes finally abandoned the satellite.
But one engineer refused to give up. It turns out that the transfer orbit paradigm above is the probably most efficient path in a single-planet system, but Earth has this anomolously large, close, Moon. And while there wasn't nearly enough fuel to get raise the perigee to geosynchronous altitude, there was more than enough fuel to raise the apogee out to lunar orbit. He was given permission to try to rescue the satellite.
In the end, two passes by the moon were made, each raising the perigee somewhat and lowering the inclination of the orbit. The remaining fuel in the satellite was used to lower the apogee back to the geosynchronous orbit altitude, but unfortunately the inclination couldn't be brought down quite to zero, so the satellite isn't in its desired orbit even today. Still, it's in an orbit where some use can be derived from it.
The satisfying conclusion to this story would be that all geosynchronous satellites are launched this way, now. Unfortunately, you can't mess with the status quo to that extent; and satellites are still, in the main, launched the old transfer-orbit way.
I wrote to the conference organizer, looking for a copy of the speech so that I could see things in context -- I can't believe that even Mundie could state things that baldly.
Remarkably, I got a reply back quite quickly from Ian Williams saying that the speech was not made available in advance, but it is begin transcribed from a recording and will be posted on the web site in a few days. He requests that we check http://www.worldcongress2002.org in the media section in a few days.
Ouch! You are correct on the math, I was lame on my explanation of the deals.
IIRC the way it works is that the 5-picture deal was renegotiated after Toy Story came out. The previous deal was a three-picture deal where Disney got the lion's share of the revenues, and the new deal is a five-picture deal with more even cost- and revenue-sharing.
So, the two pictures so far are Bug's Life and Monster's Inc; as they were released after Toy Story -- with Toy Story II not counted as it was intended to be a direct-to-video sequel.
Pixar has a five-picture deal with Disney, and for complicated reasons Toy Story II didn't count. So, there are three movies left that Pixar has to release through Disney, and that Disney has some creative control over -- after which Pixar will be able to be it's own, for better or for worse.
It appears that these three pictures are all green-lit, and are in progress at one stage or another.
I went to the World Congress web site, looking for more context to the comments, as I agree with erasmus_on above that it seemed likely that there was more to Mundie's statements than reported. Unfortunately, there is next to zero content at the site -- it appears to be more of a junket than a conference (I'll admit that this may be common.)
What I did find interesting is the last paragraph of Mundie's bio, pointing out that he was on the team at Data General that were working on the Fountainhead Project, the bad guys in Kidder's The Soul of a New Machine. This is confirmed by a Red Herring article.
One can just wonder at the FUD that was sent between the two parts of DG, as Mundie was first stretching his wings...
If blue lasers become common for DVD recorders, but the studios stick with red-laser, low-bit rate encoding for films they release, it will be possible, nay, likely, that pirates will release high-bit-rate movies that will be superior to the studio versions. That would be the ultimate insult -- with video tape, pirate copies were always worse than the originals -- with DVD they could be exactly the same, but if this decision goes through -- they could be better.
No, what was being done at the RSA conference was completely different, wonderfully practical, and available today. It's a way of cracking RSA and DES implementations in smart cards using Differential Power Analysis and Differential Fault Analysis.
There was a company at the conference that was cracking these codes by very carefully watching the power levels on the input traces to the chip. By careful analysis, you can see exactly what is happening, what numbers are being sent to the various S-Boxes in DES for instance. From this, you can determine the DES keys. They said that single DES took a couple of minutes, triple-DES only three times as long. They claimed to be able to crack RSA as well.
Differential Fault Analysis involves giving the wrong voltages to the chip, and seeing how, and when, it fails. Again, this provides clues to what numbers are being processed within the chip.
Neither of these flaws point to problems with the algorithm, but to problems with its implementation. In fact, creating inexpensive physical hardware to do secure cryptography may be impossible.
What was remarkable at the RSA conference is that most of the other booths at the trade show were selling these same smart cards, in abject denial of their flaws.
thad
NSA-sponsored Cray 4 development now makes sense.
on
Factoring Breakthrough?
·
· Score: 5, Interesting
A friend of mine worked for Cray Computer Corporation until the untimely death of Seymour Cray. The last machine they were working on was a monster, that might make more sense in terms of today's developments.
In the early nineties, CCC was working on the Cray 3, a new gallium arsenide computer. It was to have a cycle time of about 1ns (shockingly fast back then.) It was cooled by a high-pressure very high-speed mist of Flourinert suspended in helium. It was built as a series of wedges much like the Cray 1 and 2, although somewhat smaller. They built working prototype wedges, and were debugging them, while looking over their collective shoulders at the ground being gained on them by arrays of microprocessors.
One thing led to another, and it was clear that the Cray 3 would never be a commercial success. They were then given a contract to build what was called the Cray 4. The Cray 4 was a one-off machine using PIM (processor in memory) chips. These were 1-bit computers, but there were 262,144 of them in the box. The idea was that the gallium arsenide chips, wiring, and cooling system that made up the Cray 3 were just the networking system for these PIM chips, which would do the actual work.
Anyway, Cray died, and then CCC quickly died, and I don't believe that the machine was ever finished.
Slashdot Mirror
The Osprey is a spectacular concept, and the XV-15 was a completely successful prototype. It appears to me that the jump made to the production Osprey added so much complexity, and cut the tolerances so fine, that the plane is in serious trouble. Adding the folding blades, the pivoting wing, the high-pressure hydraulic system, and so forth -- it was more that could be done in one go, and the program is suffering for it.
Tilt-rotors will revolutionize aviation, there's no doubt about it. The record of the Osprey, though, really is distinctly worse than other programs that get to this stage. It's not just PR. Falsifying maintenance records didn't help, though.
I think that the Lockheed engineers wanted to build a bigger, heavier plane, and had to go to the lift fan. Direct lift on a plane as heavy as the X-35 just wasn't going to work. So, they had to try something radical. Well, they convinced some people. If it ends up working, well, great -- I'll eat my hat. Won't be the first time.
thad
I thought that the Boeing plane should have won the competition, mostly because it fulfilled the specification better; while being smaller, lighter, and immeasurably simpler. The Boeing plane didn't take off vertically, it's true, but that is also not in the specification -- it's not what was asked for. Similarly, there was no line-item for aesthetics. The Boeing direct-lift concept is the same that powers the Harrier, and is the only demonstrated successful direct lift formula. The clutch-driven lift fan is an Osprey-scale debacle waiting to happen -- mechanically clutching in 40,000 HP in a few seconds with an extremely lightweight gearbox is, I believe, untenable. They finally got it to work for a few tests, but there were a number of fairly spectacular failures along the way. The Boeing design lets the pilot shift from forward to vertical thrust and back again in a few seconds, at will, and they did it more than 100 times during the flight test program -- the Lockheed one was only clutched a handful of times.
The very wide-chord wing of the Boeing design is good for a number of structural, aerodynamic, and stealth reasons. Unfortunately, Boeing elected to change the design for the actual plane to a separate tail, rather than the delta wing -- Lockheed partisans claimed (rightly, IMHO) that this meant that the demonstrator that Boeing flew wasn't really representative of the final plane.
The one terrific thing that the Lockheed design has, the one true aerodynamic innovation, is the bump intake. There's a big bump right in front of the intakes on the Lockheed plane; it performs all of the functions of the typical intake splitter plate, purging the boundary layer, with a far more elegant, lighter, simpler, stealthier, easier-to-maintain design. Hats off to the engineers that came up with that.
I think that the Boeing design is prettier, too, but that's just me -- I'm a low-aspect ratio kind of guy.
thad
If you manage to get through the slashdotting, the story in the tecchannel web pages is amazing. The prototype Clawhammer, while limited to 800 MHz, performed shockingly well on the few, but varied, benchmarks they subjected it to. It's interesting that both Intel and AMD teach the same lesson, that MHz doesn't determine performance. Unfortunately for Intel, they demonstrate it by the P4 not running as fast as the MHz would imply, where the AMD chips run far faster than MHz would imply.
I can't wait for these chips to get out there.
thad
I know that it wasn't you people that did it, but I have to give IBM major props for that ad. For those of you that haven't seen it, there are two guys in suits watching the IBM team play basketball. There's "Middleware", "Applications", and, of course, "Linux", a tall, lumbering, bear of a basketball player.
One guy says to the other, "You know, that guy Linux doesn't get paid a cent for this"
and the other guy says "No! Why does he do it then?"
And the first guy says "Loves the game..."
Well, that just about says it all. Thanks IBM.
thad
From the first hundred pages or so, it doesn't seem terribly revolutionary. I read the stunning
Cellular Automata Machines by Toffoli and Margolus back when it was first issued in the late eighties, and it seemed far more exciting, and far less egomaniacal.
But, I'll continue to plow on, and see what develops. From the great review above, it does appear that it gets better soon.
Read Cellular Automata Machines, though. It's completely awesome.
thad
Back in 1988, Graham Walters and I, at PDI, built a system in collaboration with Jim Henson and his creature shop, and with Kirk Thatcher (who since went on to greater things at Henson), to build the 'Waldo' character for The Jim Henson Hour. The idea was that this character would be controlled by a waldo, which would sense the position, orientation, and mouth angles, and display the character, blue-screen keyed onto the image on the screen in real time.
The beauty of this system is that Henson puppeteers always work by watching their images on TV monitors, so this kind of digital character wasn't even second-nature to them -- it was exactly how they'd been performing characters all along. Among the nice things about Waldo is that he didn't have to hang out at the bottom of the screen with all of the rest of the other puppets. This system was implemented and run on an old Power Series SGI borrowed from Sheridan College.
I saw Davey Goelz (Gonzo, and others) at Siggraph this year, at the Henson booth on the show floor. They're selling a somewhat improved version of the same waldo mechanism that we used 14 years ago. Davey got us out of a jam on the first Henson Hour show, as somehow we lost the mouth-opening information from the tracks that Henson recorded on the set. Davey lived right near us at PDI, and came down and laid those back in, mimicing Jim's style. I don't think that he ever found our, and it's tragically too late now.
Anyway, The Jim Henson Hour was too good, and perhaps a little too different, for American TV, and only 12 episodes were ever made.
thad
Two companies that have been pioneers in UWB are Pulson and Aetherwire.
Pulson (and its predecessor company, Time Domain) has been desparately trying to commercialize this technology for radio communication for years. More than five years ago they demonstrated a few-milliwatt UWB radio with 100-mile range. They have mostly been held back by patents taken out by Lawrence Livermore. Livermore claims to have invented all of this stuff, and has been rediculously rough on licensing. Also, the FCC has been unclear until very recently on how it would license UWB.
Aetherwire has attempting to use UWB technology to build localizers, basically extremely short range, extremely low-power peer-to-peer short-range version of GPS. The localizers would all cooperate at keeping track of where the other ones were within a few hundred meter radius. If you've read A Deepness in the Sky by Vernor Vinge, it's all about localizers.
Now that the FCC has cleared the way, I expect to see tremendous progress in UWB. It's going to revolutionize many fields, from radio to positioning to radar.
thad
I know that I'm adding this comment far to late to get noticed, but it must be said. I'm certain that the only reason that MS is making this announcment, and certainly the reason that they are making it now is to impress Colleen Kollar-Kotelly -- saying that they are not trying to take over the world, after all.
thad
One of the most important space assets is a global positioning system. So far, the US has allowed free use of our DoD-sponsored GPS, but everybody knows that policy could change at some time in the future. The Europeans felt that without their own system, their soverignty was threatened, so they proposed their own system, called Galileo [must have been some Europoean guy.]
Anyway, the US is all in a sweat about this, and has put quite a bit if pressure on the Europeans to get them to forego this system. The US says that it's not needed, that it would interfere with our system, that it would destabilize the planet -- basically the entire bag of boogeymen. To it's credit, Europe has recently reversed course and decided that they would create this system on their own. Germany had been leaning toward the US position, but just changed their mind.
Galileo is not very well defined yet (even basic things like how many satellites, and which orbit configuration), but most people expect that it will be somewhat better than the current GPS system -- although the US insists that it's super-duper GPSII system will be better than Galileo, whatever Galileo ends up being.
thad
LCD's really are dramatically better than they used to be. Mike Massee's statement above was certainly true a few years ago, but it is no longer the case. At Illusion Arts, many of the matte painters are using LCD screens. Some of these people went straight from real paint to LCDs without ever seeing CRTs.
thad
The last time I bought an ATI card I was shocked by the fact that the Linux driver had incredibly bad 2D performance. ATI had paid the Precision Insight folks to write the 3D driver, but didn't include any money for accelerated 2D performance.
I don't know if this has been remedied since then, this was the status 6 months ago.
thad
We live about 100 yards from the Mulholland Drive mentioned at the end of the article, right in the middle of Los Angeles. Interestingly, and frustratingly, even though this part of Bel Air and Beverly Hills is full of people who would desparately like to have broadband access, there is none. No DSL, no cable modems. ATT has pulled out its fixed-wireless system. Metricom of course went belly-up.
Adelphia would be our cable modem provider. They've been busily laying cable for the last year, and have all but completed their network. Now I read in this story, Adelpha claims that it being crippled by DWP, because they can't get power to their network.
I wonder if the Department of Water and Power sees Adelphia as competition, and is inhibiting them in the obvious way. Or, this might be another case where you shouldn't attribute to malice what can equally be explained as bumbling by a cable company.
It will be interesting. Adelphia claims that they'll light up the fibers here within the next month or so. I can't wait.
thad
This post will try to make Showtime qualify for the 'News for Nerds' motto of Slashdot.
You might not think of this as a big FX movie, after all, it's just a cop buddy movie, right? Well, it turns out that my little company did over 140 digital FX shots for this film. This demonstrates among other things that computer graphics effects are a big part of almost any film these days -- it gives the director and writers freedom to shoot things more effectively, and allows the director and writer the freedom to improve the film in the post-production phase.
I'll try not to reveal too many plot points to those of you who haven't seen the film -- but if you don't want to know anything about the film you can stop reading right here.
Ok. The biggest thing that we did was build the environment outside the penthouse -- buildings, reflections, and helicopters. The challenge was to keep that environment alive (without the cliche flocks of pigeons.) We did this mostly by observing that windows in skyscrapers flex somewhat in the wind, so that the distortions of the reflections in the windows are always changing. There are a few places that you can see near the ground, and we added little sparkles of light from car windshields, things like that.
We also added a bunch of lasers to the guns, and a bunch of sparks in the gun show sequence. Finally, we did all the 'videoization' and changed the license plate of Eddie Murphy's car for some shots. There were a few dozen other little things here and there, but that's the majority of the shots. By the way, in reality there is a heliport on the roof of the Westin Bonaventure in downtown
Los Angeles.
All of this work was done by a team of 10 or so people over about six months. We used a combination of SGI and Linux boxes to do the animation design, and our render garden (it's too small to be a proper render farm) of Linux boxes to do all of the batch rendering.
thad
I've told my employees that if I ever miss a paycheck, they should quit immediately, regardless of what I say at the time. I tell them not to believe me if I say it will get better. I've never seen things get better for a company once they stop being able to make payroll. I don't suppose that the visual effects business is any different than any other business in that respect. Once you start digging a hole it becomes increasingly impossible to ever get out of it.
The problem is, that once a company starts foundering, the founders often begin to lose touch with reality and start making promises that they can't keep -- whether or not they know that is not really an issue. The hole is not only financial (although that's a big enough hurdle on its own) but it's also bad will, that is, the accumulated acrimony festering within the company.
thad
I have been gently pressing the organizers of the 'world conference' about getting the original copy of Mundie's speech, but so far I have been unable to get it. They claimed, at first, to be transcribing the speech, and that it would be available at the media part of the site, but so far it hasn't appeared.
I really want to see the original source, as I believe that it's quite likely that Mundie's reported words are not particularly accurate, and they are surely quoted out of contest. I'm most interested, of course, because I think that the original text is, if anything, more strident and open for redicule.
I've got an email log of my conversations with the World Conference orgranzers that I'd be willing to share with anybody, on request, just send an email. Perhaps with a few more people asking we can get the transcript.
thad
I love cryptome, I read it every day. That said, the report John Young has put up is the unclassified version; although he's trolling for the classified version.
Hey Taco, any idea when the results of the John Young questions will be posted? It's been six months or so...
thad
About five years ago, Hughes launched a communication satellite. To be useful, these have to reach a geosynchronous orbit. It turns out that the most efficient way to get to from a low-earth orbit to a geosynchronous orbit is to fire a rocket twice [or so it was thought.] The first firing raises the apogee of the orbit to the geosynchronous altitude; so that the satellite is in a very elliptical orbit. Then, when the satellite is at apogee, you fire the rocket again to circularize the orbit. Usually this same motor changes the plane of the orbit as well. Most satellites are launched into orbits inclined to the equator somewhat, and geosynchronous satellites have to be over the equator. It is most efficient to make the orbital plane change at apogee, too.
There was a rash of apogee kick motor failures, and in this particular satellite the motor failed, leaving the satellite in a uselessly ellipitical orbit. There were small thrusters on the satellite which were to be used for station-keeping (small orbital adjustments) but it didn't have nearly enough propellant to raise the perigee. Hughes finally abandoned the satellite.
But one engineer refused to give up. It turns out that the transfer orbit paradigm above is the probably most efficient path in a single-planet system, but Earth has this anomolously large, close, Moon. And while there wasn't nearly enough fuel to get raise the perigee to geosynchronous altitude, there was more than enough fuel to raise the apogee out to lunar orbit. He was given permission to try to rescue the satellite.
In the end, two passes by the moon were made, each raising the perigee somewhat and lowering the inclination of the orbit. The remaining fuel in the satellite was used to lower the apogee back to the geosynchronous orbit altitude, but unfortunately the inclination couldn't be brought down quite to zero, so the satellite isn't in its desired orbit even today. Still, it's in an orbit where some use can be derived from it.
The satisfying conclusion to this story would be that all geosynchronous satellites are launched this way, now. Unfortunately, you can't mess with the status quo to that extent; and satellites are still, in the main, launched the old transfer-orbit way.
thad
I wrote to the conference organizer, looking for a copy of the speech so that I could see things in context -- I can't believe that even Mundie could state things that baldly.
Remarkably, I got a reply back quite quickly from Ian Williams saying that the speech was not made available in advance, but it is begin transcribed from a recording and will be posted on the web site in a few days. He requests that we check http://www.worldcongress2002.org in the media section in a few days.
thad
Ouch! You are correct on the math, I was lame on my explanation of the deals.
IIRC the way it works is that the 5-picture deal was renegotiated after Toy Story came out. The previous deal was a three-picture deal where Disney got the lion's share of the revenues, and the new deal is a five-picture deal with more even cost- and revenue-sharing.
So, the two pictures so far are Bug's Life and Monster's Inc; as they were released after Toy Story -- with Toy Story II not counted as it was intended to be a direct-to-video sequel.
thad
Pixar has a five-picture deal with Disney, and for complicated reasons Toy Story II didn't count. So, there are three movies left that Pixar has to release through Disney, and that Disney has some creative control over -- after which Pixar will be able to be it's own, for better or for worse.
It appears that these three pictures are all green-lit, and are in progress at one stage or another.
thad
I went to the World Congress web site, looking for more context to the comments, as I agree with erasmus_on above that it seemed likely that there was more to Mundie's statements than reported. Unfortunately, there is next to zero content at the site -- it appears to be more of a junket than a conference (I'll admit that this may be common.)
What I did find interesting is the last paragraph of Mundie's bio, pointing out that he was on the team at Data General that were working on the Fountainhead Project, the bad guys in Kidder's The Soul of a New Machine. This is confirmed by a Red Herring article.
One can just wonder at the FUD that was sent between the two parts of DG, as Mundie was first stretching his wings...
thad
If blue lasers become common for DVD recorders, but the studios stick with red-laser, low-bit rate encoding for films they release, it will be possible, nay, likely, that pirates will release high-bit-rate movies that will be superior to the studio versions. That would be the ultimate insult -- with video tape, pirate copies were always worse than the originals -- with DVD they could be exactly the same, but if this decision goes through -- they could be better.
Interesting times.
thad
In my letter my name is spelled correctly as 'Beier' but it is incorrectly spelled in the alphabetical listing as 'Bejer'.
What is odd is that the only previous time that I've seen that mispelling is in the credits to the movie Showgirls. Kinda makes you wonder.
Thad Beier
No, what was being done at the RSA conference was completely different, wonderfully practical, and available today. It's a way of cracking RSA and DES implementations in smart cards using Differential Power Analysis and Differential Fault Analysis.
There was a company at the conference that was cracking these codes by very carefully watching the power levels on the input traces to the chip. By careful analysis, you can see exactly what is happening, what numbers are being sent to the various S-Boxes in DES for instance. From this, you can determine the DES keys. They said that single DES took a couple of minutes, triple-DES only three times as long. They claimed to be able to crack RSA as well.
Differential Fault Analysis involves giving the wrong voltages to the chip, and seeing how, and when, it fails. Again, this provides clues to what numbers are being processed within the chip.
Neither of these flaws point to problems with the algorithm, but to problems with its implementation. In fact, creating inexpensive physical hardware to do secure cryptography may be impossible.
What was remarkable at the RSA conference is that most of the other booths at the trade show were selling these same smart cards, in abject denial of their flaws.
thad
A friend of mine worked for Cray Computer Corporation until the untimely death of Seymour Cray. The last machine they were working on was a monster, that might make more sense in terms of today's developments.
In the early nineties, CCC was working on the Cray 3, a new gallium arsenide computer. It was to have a cycle time of about 1ns (shockingly fast back then.) It was cooled by a high-pressure very high-speed mist of Flourinert suspended in helium. It was built as a series of wedges much like the Cray 1 and 2, although somewhat smaller. They built working prototype wedges, and were debugging them, while looking over their collective shoulders at the ground being gained on them by arrays of microprocessors.
One thing led to another, and it was clear that the Cray 3 would never be a commercial success. They were then given a contract to build what was called the Cray 4. The Cray 4 was a one-off machine using PIM (processor in memory) chips. These were 1-bit computers, but there were 262,144 of them in the box. The idea was that the gallium arsenide chips, wiring, and cooling system that made up the Cray 3 were just the networking system for these PIM chips, which would do the actual work.
Anyway, Cray died, and then CCC quickly died, and I don't believe that the machine was ever finished.
thad