This is US vs. Kilbride, decided last October. It apparently took Sedo a few months to notice.
It's actually a porno spam case left over from the Bush Administration. It's not like the Justice Department was doing anything effective about spam in general.
There's a difference between devices that are mostly for receiving information, and those that are for doing something with it. Music players, "e-book" readers, navigation devices, and entertainment devices in general are mostly-receive. They need a much simpler interface than a creation device. Try to cram a CAD application into the iPhone interface. It's possible, but it's not happy there.
This is a bigger distinction than the form factor. Mostly-receive devices can get along with a blunt interface of big buttons.
If anything, the Linux kernel changes too much. It ought to settle down into a tight little kernel that's changed only for rare bug fixes. The "monolithic kernel" concept has gotten somewhat out of hand. Arguably, no USB device driver or printer driver should be in the kernel or have any significant privileges. That alone would cut way down on kernel mods.
Does any major software still need the 16-bit subsystem?
Amusingly, when I first installed Windows NT 3.51, back around 1996, the 16-bit subsystem was optional, like the OS/2 subsystem, and I had it turned off. Everything worked fine. In NT 4, they let the kode kiddies from the Windows 95 group put legacy code into NT, some of which still ran in 16-bit mode, and the 16-bit subsystem was always on.
The gamer blog has it wrong, the article poster didn't help, and the Slashdot "editors" blew it as usual.
Read the article in JoongAng Daily (which they offer in English). The key issue here is that online gambling is illegal in Korea, and two game players were charged criminally for making money from an online game. The Supreme Court of Korea ruled that they were not gambling, so they don't get fined.
This decision doesn't affect relationships between players and game operators. It's not about EULA enforceability or property rights. It's a criminal law issue. If you trade game currency, you're not going to be fined or go to jail in Korea. Whether a game site can ban you is a separate issue.
Somewhere in Mountain View, servers are now analyzing all activity of Google's employees in China. And their friend. And their friend's friends. And the people they email. And everyone who got in range of a Google security camera. And all the their friends. And the people they email.
I'd thought this was a video game problem, with focus on the screen narrowing awareness to a narrow cone.
I own a horse, and horse barns tend to have teenagers around. The kids who ride usually have good situational awareness. Part of riding training is to ride in a busy ring, constantly aware of what all the other horses are doing. But often the parents drag along non-riding kids. When they're using their iPhone or Nintendo DS, they're totally oblivious.
Even offline, they're not aware of what's going on behind them. I've had the experience more than once of riding up behind a child and having to work at getting their attention just to get them out of the way. I'm not talking about not noticing something at a distance. There are kids that don't notice the horse's nose directly above their head. This is a lack of basic survival skills. What would these kids do in a bad neighborhood?
The first time this happened, about five years ago, I thought that perhaps someone had brought a retarded kid to the barn. But I've seen this too many times since then.
If you're bringing up kids, make sure they do something that forces wide situational awareness. Riding. Soccer. Martial arts. Paintball. Birdwatching. Otherwise you're bringing up a victim.
The article has too much hype, but the actual work has some potential.
For the limited problem they're really addressing, extracting certain data about sports teams and corporate mergers, this approach might work.
Both of those areas have the property that you can get structured data feeds on the subject. Bloomberg will sell you access to databases which report mergers in a machine-processable way; some stock analysis programs need that data. Sports statistics are, of course, available on line. So the program's extraction of that info from news stories intended for humans can be checked. This allows supervised learning. The program can tell what it got right and what it got wrong.
When they can distinguish between a merger that's being talked about, one which entered negotiations but was not completed, one which went for DoJ approval and was rejected, and one which was completed, they'll have something. Until then, they're probably won't outperform "'merger' NEAR 'companyname'" queries.
It's amusing that we don't have "high end" computers for multimedia use. Features might include:
No firmware runs in System Management Mode, stealing cycles from the main CPU.
No paging.
Operating system is tested and certified for interrupt response under 1us, 100% of the time. (Hard real-time operating systems like QNX can do this. Linux and Windows still have excessive interrupt lockout times; I think Linux is now below 1ms if you don't have any crappy drivers installed, but 1us is a long way off.)
Support for "sporadic scheduling", where the OS guarantees, say, 20% of the CPU every 1ms to a task that requests it. This allows playing multimedia with no breaks while doing something else. If you try to open too many multimedia windows, the scheduling request is rejected, because you're out of capacity.
Disk access prioritization, so that CPU priority affects disk access priority. (QNX has this).
All solid state disks.
These are the kind of specs you see in hard real time systems that have to run both time-critical and non-time-critical code. "Multimedia PCs" ought to have specs like that, but they don't. So you still get pausing and stuttering if something else interferes with playback.
A typical test in the real time world is to hook up a square wave generator to an input pin and a digital oscilloscope to an output pin. You then run a program which is waiting for interrupts triggered by the input pin, and when the user process triggered by the interrupt gets control, it turns on the output pin. You load up the CPU with other, lower-priority tasks. You watch the results on a storage 'scope, timing the time from input to output. You expect all the spikes to be below the promised time threshold. If there are any outliers, users get annoyed, file bug reports, and it gets fixed. This is how you get rid of "jitter" at the OS level.
This, as the article points out, is basically a beefed up version of magnetic forming. Magnetic forming has been around for decades. It's useful mostly for compressing cylindrical objects, so it's used on couplings, tube joints, and similar round objects I've seen it used in making hydraulic spool valves. It's a way to apply a completely symmetric radial squeezing force, which is hard to do at high precision with stamping dies or presses.
Here are some examples of parts formed by magnetic forming.
But for punching holes, there's no obvious advantage to magnetic forming.
The US military wants robots. More robots. Robots that kill. Now.
Read Failure To Field The Right Robots Costs Lives, General Says. Lt. General Rick Lynch, commander of the U.S. Army's 3rd Armored Corps, wants autonomous killbots. His corps lost 155 soldiers in Iraq, and he claims that 80% of them would have been saved if the right kind of robots were deployed. On watching "hotspots" for enemy activity: "Robots can take the soldiers' places. They can continuously keep watch on an area, and if nefarious activity is spotted, we can take appropriate action.... We can kill those bastards before they plant the IEDs"
This is a combat general in charge of a major Army command making it happen.
I have no problem with this. There was a reasonable likelihood of a megadeath sized pandemic. Appropriate steps were taken to prevent it. Some of those steps may have been unnecessary, but it didn't hurt and wasn't outrageously expensive.
The swine flue vaccination campaign in the US probably has already saved more lives than the entire Department of Homeland Security.
Imagine them communicating with each other in pack and relaying GPS location data. If one finds a target, they start to zero in on the victim.
Reality: DARPA funded work on that in 1997. Sandia made it work. The Sandia concept turned out not to be too useful militarily, but paved the way for the
Precision Urban Hopper..
This is embarrassing. Embarrassing for Thompson, which owns the RCA brand. Embarrassing for press who took it seriously. Serious career trouble for whomever authorized licensing the name to the clowns behind this.
We'll probably hear some disclaimer from Thompson shortly.
Another "slave to the server" DRM scheme. Those have a finite lifetime.
What's the longest-lived "slaved to a server" DRM scheme? Has any such scheme been working for ten years? iTunes may be the oldest, but they didn't support video until 2005, and they've been moving away from DRM on audio.
Think of what al-Queda could do with the signing key for Windows Update.
No, the difference is that there's now a whole world of stuff and people out there that will respond to them instantly. Until recently, you could want that, but unless you were rich enough to have servants and had parents who let you give them orders, you couldn't get that.
This is a step up from five hours a day passively watching TV.
You realize in their demo video, they write a adblock like jetpack with 80 lines of code.
A demo of undocumented features, perhaps? The manual says those features aren't implemented yet. If the Mozilla crowd wants developers to use their new API, they'll have to document it.
That leads to Jetpack Extension Proposal #17 - Page Mods, which discusses how to implement Greasemonkey-like functionality using Jetpack. Current status is "Implementing (since May 27, 2009)".
It's significant that Mozilla gave priority to implementing "themes" and such, which are needed for vendor-branded browsers, while putting off implementation of user-oriented features like ad blocking. Is this a back-door effort to get ad-blocking out of Firefox?
A touch-screen in a car, at least for the driver, is a terrible idea. It can't be operated by feel; the driver has to look away from the road, and probably for more than a second. Not good. Twittering while driving? Please. "Fully Loaded", a Bruce McCall drawing, isn't a design goal.
Auto designers, desperately trying to get margins up with "more car per car" (an old GM slogan) are hanging on unneeded features that are cheap to install. Overpriced car stereos aren't enough any more. Giant hood ornaments are out (there's a "pedestrian impalement" test cars have to pass, in response to a period in the 1950s when auto hoods were weaponized). So now we have dashboard gimmicks.
In aviation, this is called the "head-down time" problem, and efforts are made to minimize head-down time. The military takes this to an extreme in fighters, with the HOTAS ("Hands On Throttle and Stick") concept. This leads to a proliferation of buttons on the throttle and stick, though. Aviation people think hard about how many seconds of head-down time it takes to do something.
If you want to cause accidents, put in a touch screen that's stateful, so the driver has to look. Then give it a timeout, so it goes back to the ground state if the driver doesn't give it undivided attention. This forces the driver to look away from the road. One of the examples in the original article looks very like that.
23andme, the commercial DNA analysis service, checks 580,000 sites in DNA.
23andme probably has enough data to validate the quality of the FBI's marker selection. That's a good way to check. Identical twins do match, even at the 23andme level of analysis.
My experience is that Google isn't geared toward customer service and it seems like they could care less.
Right. Google has never before offered a product that requires a substantial customer service organization. The one major Google service that needs customer support, Google business e-mail, they support by sending their customers to third party "solution providers".
It's surprisingly hard to automate page-turning. I saw the first page-turning machine many years ago, at the Census Bureau. It was used for 1970 Census form booklets, and used a vacuum belt to hold the booklet down while a wheel with vacuum holes rolled over the page to turn the page. This only worked for booklets with known dimensions, and it was rather rough on the booklets. But it was fast, doing about two flips a second.
It's such a boring job for humans that they screw up. A hand appears in the picture, or they turn two pages. So you need automation, or at least automated error checking.
The problem with mechanism design is making it both fast and gentle. There are lots of things that will work at one page every five seconds. Getting to two pages a second and never tearing one is tough. Most of the existing designs are simplistic; they're just some dumb mechanism making a repetitive motion with an air picker. The book-scanning developers haven't progressed to closed-loop force control yet.
Festo, the German robotics and actuator company, could probably build a better page turner. They build a wide range of machines which handle delicate objects fast in production environments. Their Bionic Tripod with Fin-Gripper is an example.
Or Gmail? Free email accounts are spammer magnets. Google doesn't even try hard to stop Gmail Account Creator ("For when one email account isn't enough.")
Mail from a Hotmail account just screams "loser". That thing should just die a quiet death, like GeoCities.
Slashdot Mirror
This is US vs. Kilbride, decided last October. It apparently took Sedo a few months to notice.
It's actually a porno spam case left over from the Bush Administration. It's not like the Justice Department was doing anything effective about spam in general.
There's a difference between devices that are mostly for receiving information, and those that are for doing something with it. Music players, "e-book" readers, navigation devices, and entertainment devices in general are mostly-receive. They need a much simpler interface than a creation device. Try to cram a CAD application into the iPhone interface. It's possible, but it's not happy there.
This is a bigger distinction than the form factor. Mostly-receive devices can get along with a blunt interface of big buttons.
If anything, the Linux kernel changes too much. It ought to settle down into a tight little kernel that's changed only for rare bug fixes. The "monolithic kernel" concept has gotten somewhat out of hand. Arguably, no USB device driver or printer driver should be in the kernel or have any significant privileges. That alone would cut way down on kernel mods.
Does any major software still need the 16-bit subsystem?
Amusingly, when I first installed Windows NT 3.51, back around 1996, the 16-bit subsystem was optional, like the OS/2 subsystem, and I had it turned off. Everything worked fine. In NT 4, they let the kode kiddies from the Windows 95 group put legacy code into NT, some of which still ran in 16-bit mode, and the 16-bit subsystem was always on.
The gamer blog has it wrong, the article poster didn't help, and the Slashdot "editors" blew it as usual. Read the article in JoongAng Daily (which they offer in English). The key issue here is that online gambling is illegal in Korea, and two game players were charged criminally for making money from an online game. The Supreme Court of Korea ruled that they were not gambling, so they don't get fined.
This decision doesn't affect relationships between players and game operators. It's not about EULA enforceability or property rights. It's a criminal law issue. If you trade game currency, you're not going to be fined or go to jail in Korea. Whether a game site can ban you is a separate issue.
Somewhere in Mountain View, servers are now analyzing all activity of Google's employees in China. And their friend. And their friend's friends. And the people they email. And everyone who got in range of a Google security camera. And all the their friends. And the people they email.
I'd thought this was a video game problem, with focus on the screen narrowing awareness to a narrow cone.
I own a horse, and horse barns tend to have teenagers around. The kids who ride usually have good situational awareness. Part of riding training is to ride in a busy ring, constantly aware of what all the other horses are doing. But often the parents drag along non-riding kids. When they're using their iPhone or Nintendo DS, they're totally oblivious.
Even offline, they're not aware of what's going on behind them. I've had the experience more than once of riding up behind a child and having to work at getting their attention just to get them out of the way. I'm not talking about not noticing something at a distance. There are kids that don't notice the horse's nose directly above their head. This is a lack of basic survival skills. What would these kids do in a bad neighborhood?
The first time this happened, about five years ago, I thought that perhaps someone had brought a retarded kid to the barn. But I've seen this too many times since then.
If you're bringing up kids, make sure they do something that forces wide situational awareness. Riding. Soccer. Martial arts. Paintball. Birdwatching. Otherwise you're bringing up a victim.
The article has too much hype, but the actual work has some potential. For the limited problem they're really addressing, extracting certain data about sports teams and corporate mergers, this approach might work.
Both of those areas have the property that you can get structured data feeds on the subject. Bloomberg will sell you access to databases which report mergers in a machine-processable way; some stock analysis programs need that data. Sports statistics are, of course, available on line. So the program's extraction of that info from news stories intended for humans can be checked. This allows supervised learning. The program can tell what it got right and what it got wrong.
When they can distinguish between a merger that's being talked about, one which entered negotiations but was not completed, one which went for DoJ approval and was rejected, and one which was completed, they'll have something. Until then, they're probably won't outperform "'merger' NEAR 'companyname'" queries.
It's amusing that we don't have "high end" computers for multimedia use. Features might include:
These are the kind of specs you see in hard real time systems that have to run both time-critical and non-time-critical code. "Multimedia PCs" ought to have specs like that, but they don't. So you still get pausing and stuttering if something else interferes with playback.
A typical test in the real time world is to hook up a square wave generator to an input pin and a digital oscilloscope to an output pin. You then run a program which is waiting for interrupts triggered by the input pin, and when the user process triggered by the interrupt gets control, it turns on the output pin. You load up the CPU with other, lower-priority tasks. You watch the results on a storage 'scope, timing the time from input to output. You expect all the spikes to be below the promised time threshold. If there are any outliers, users get annoyed, file bug reports, and it gets fixed. This is how you get rid of "jitter" at the OS level.
This, as the article points out, is basically a beefed up version of magnetic forming. Magnetic forming has been around for decades. It's useful mostly for compressing cylindrical objects, so it's used on couplings, tube joints, and similar round objects I've seen it used in making hydraulic spool valves. It's a way to apply a completely symmetric radial squeezing force, which is hard to do at high precision with stamping dies or presses. Here are some examples of parts formed by magnetic forming.
But for punching holes, there's no obvious advantage to magnetic forming.
The US military wants robots. More robots. Robots that kill. Now.
Read Failure To Field The Right Robots Costs Lives, General Says. Lt. General Rick Lynch, commander of the U.S. Army's 3rd Armored Corps, wants autonomous killbots. His corps lost 155 soldiers in Iraq, and he claims that 80% of them would have been saved if the right kind of robots were deployed. On watching "hotspots" for enemy activity: "Robots can take the soldiers' places. They can continuously keep watch on an area, and if nefarious activity is spotted, we can take appropriate action. ... We can kill those bastards before they plant the IEDs"
This is a combat general in charge of a major Army command making it happen.
I have no problem with this. There was a reasonable likelihood of a megadeath sized pandemic. Appropriate steps were taken to prevent it. Some of those steps may have been unnecessary, but it didn't hurt and wasn't outrageously expensive.
The swine flue vaccination campaign in the US probably has already saved more lives than the entire Department of Homeland Security.
Imagine them communicating with each other in pack and relaying GPS location data. If one finds a target, they start to zero in on the victim.
Reality: DARPA funded work on that in 1997. Sandia made it work. The Sandia concept turned out not to be too useful militarily, but paved the way for the Precision Urban Hopper..
This is embarrassing. Embarrassing for Thompson, which owns the RCA brand. Embarrassing for press who took it seriously. Serious career trouble for whomever authorized licensing the name to the clowns behind this.
We'll probably hear some disclaimer from Thompson shortly.
Another "slave to the server" DRM scheme. Those have a finite lifetime.
What's the longest-lived "slaved to a server" DRM scheme? Has any such scheme been working for ten years? iTunes may be the oldest, but they didn't support video until 2005, and they've been moving away from DRM on audio.
Think of what al-Queda could do with the signing key for Windows Update.
Some lame e-mail domains:
Sounds like how teenagers have always been.
No, the difference is that there's now a whole world of stuff and people out there that will respond to them instantly. Until recently, you could want that, but unless you were rich enough to have servants and had parents who let you give them orders, you couldn't get that.
This is a step up from five hours a day passively watching TV.
You realize in their demo video, they write a adblock like jetpack with 80 lines of code.
A demo of undocumented features, perhaps? The manual says those features aren't implemented yet. If the Mozilla crowd wants developers to use their new API, they'll have to document it.
Right now, it looks like AdBlock, Flashblock, CustomizeGoogle, and my own AdRater couldn't be implemented under JetPack. The Jetpack API documentation has a section "Content - Methods for interacting with web pages. That's the mechanism anything that deals with ads needs. That leads to "Page modifications", which leads to This documentation is under development. Please see the page modifications API proposal for now."
That leads to Jetpack Extension Proposal #17 - Page Mods, which discusses how to implement Greasemonkey-like functionality using Jetpack. Current status is "Implementing (since May 27, 2009)".
So the functionality needed for AdBlock, etc. is vaporware. It's not even clear that, if implemented, the proposed mechanism would support AdBlock. The author of Adblock Plus wrote last month "Jetpack has to support Adblock Plus, not the other way around. As it is now, Jetpack isn't suitable for complicated extensions."
It's significant that Mozilla gave priority to implementing "themes" and such, which are needed for vendor-branded browsers, while putting off implementation of user-oriented features like ad blocking. Is this a back-door effort to get ad-blocking out of Firefox?
A touch-screen in a car, at least for the driver, is a terrible idea. It can't be operated by feel; the driver has to look away from the road, and probably for more than a second. Not good. Twittering while driving? Please. "Fully Loaded", a Bruce McCall drawing, isn't a design goal.
Auto designers, desperately trying to get margins up with "more car per car" (an old GM slogan) are hanging on unneeded features that are cheap to install. Overpriced car stereos aren't enough any more. Giant hood ornaments are out (there's a "pedestrian impalement" test cars have to pass, in response to a period in the 1950s when auto hoods were weaponized). So now we have dashboard gimmicks.
In aviation, this is called the "head-down time" problem, and efforts are made to minimize head-down time. The military takes this to an extreme in fighters, with the HOTAS ("Hands On Throttle and Stick") concept. This leads to a proliferation of buttons on the throttle and stick, though. Aviation people think hard about how many seconds of head-down time it takes to do something.
If you want to cause accidents, put in a touch screen that's stateful, so the driver has to look. Then give it a timeout, so it goes back to the ground state if the driver doesn't give it undivided attention. This forces the driver to look away from the road. One of the examples in the original article looks very like that.
The FBI's database only uses 15 markers, checking 15 sites in DNA. That's not good enough, and there are false matches. The problem is that they're using DNA technology from about 1990.
23andme, the commercial DNA analysis service, checks 580,000 sites in DNA. 23andme probably has enough data to validate the quality of the FBI's marker selection. That's a good way to check. Identical twins do match, even at the 23andme level of analysis.
My experience is that Google isn't geared toward customer service and it seems like they could care less.
Right. Google has never before offered a product that requires a substantial customer service organization. The one major Google service that needs customer support, Google business e-mail, they support by sending their customers to third party "solution providers".
We finally get a display technology with zero flicker, the LCD, and the 3D crowd has to put it back. Yuck.
It's surprisingly hard to automate page-turning. I saw the first page-turning machine many years ago, at the Census Bureau. It was used for 1970 Census form booklets, and used a vacuum belt to hold the booklet down while a wheel with vacuum holes rolled over the page to turn the page. This only worked for booklets with known dimensions, and it was rather rough on the booklets. But it was fast, doing about two flips a second.
It's such a boring job for humans that they screw up. A hand appears in the picture, or they turn two pages. So you need automation, or at least automated error checking.
The problem with mechanism design is making it both fast and gentle. There are lots of things that will work at one page every five seconds. Getting to two pages a second and never tearing one is tough. Most of the existing designs are simplistic; they're just some dumb mechanism making a repetitive motion with an air picker. The book-scanning developers haven't progressed to closed-loop force control yet.
Festo, the German robotics and actuator company, could probably build a better page turner. They build a wide range of machines which handle delicate objects fast in production environments. Their Bionic Tripod with Fin-Gripper is an example.
Or Gmail? Free email accounts are spammer magnets. Google doesn't even try hard to stop Gmail Account Creator ("For when one email account isn't enough.") Mail from a Hotmail account just screams "loser". That thing should just die a quiet death, like GeoCities.