A good way to find out if a site is selling something is by looking for links that lead to forms that take credit cards.
That's a good spam-filtering algorithm, too.
As I keep telling people who fight spam, "follow the money". Quit worrying about where the spam is coming from. Follow where the money goes.
"Is this web site selling something"?
on
IBM vs. Content Chaos
·
· Score: 3, Insightful
Search engine spiders need to understand more about sites. Things like this:
The site is selling something.
The page is composed of multiple unrelated articles or ads, each one of which should be viewed as a separate entity for search purposes.
The page is part of a blog.
Content on this site duplicates that found on other sites.
The site is owned by an organization with a known Dun and Bradstreet number. (If a site is selling something, and its Whois info doesn't match the DNB corporation database, it should be downgraded in search position. This would encourage honest Whois info.)
DARPA used to have a huge role in CS research. But that's no longer the case.
The defining moment was when they pulled the plug in Berkeley's BSD group in the early 1980s (DARPA decided to fund Mach instead) and BSD went on anyway, with private funding.
I still remember the transition from 8-bit to 16-bit IMP numbers. And the transition to 32-bit IP addresses.
In retrospect, Xerox had it right in XNS - 48-bit MAP addresses on the LAN, and 48-bit net numbers for routing between LANs. When the transition to IP came along, the old ARPANET lobby wanted to just transition by putting their IMP number in the second half of the IP address, and adding [10.0.xxx.xxx].
That's how we got into this mess of class A, B and C networks, netblocks, NAT, and all this other junk.
IPv6 is in some ways worse, because the interpretation of those 128 bits is complicated. Not everybody gets an autonomous system number and gets to participate in routing.
Basically, they're selling money with credit cards. Credit card merchant accounts don't normally allow you to do that. That's usually considered a cash advance, or loan, for which you have to be a lending regulated institution.
Bitpass disclaims all possible liability for transactions. Then they state that transactions are anonymous to earners. This makes refunds almost impossible.
They use "Entrust" SSL certificates, which Entrust says guarantee almost nothing. Verisign at least does some minimal verification.
Bitpass claims that they're not a merchant, they're not a bank, and they're not an intermediary, in further attempts to avoid liability. Contrast this with iBill, which acts as a reseller and offers customer service and refunds. Bitpass doesn't do that.
They want 90 days to refund your money if you cancel.
The disclosures required by California law,
the name and actual business address of the business, don't appear before the "enter credit card" page.
The info is on the site, but that's not good enough.
Bitpass is either a depository institution, like a bank, or a "check seller", like companies that sell money orders. Either way, they should have the appropriate license and be subject to bank regulations and audits. You can't just go into business handling other people's money without oversight.
ISPs can file lawsuits under this act. Have any been filed? Why not?
I want to subscribe to a service which sues spammers. The CAN-SPAM act's definition of an ISP seems to include a service like SpamCop. But SpamCop doesn't have a litigation staff, and their parent, IronPort sells spammer-friendly million-email-per-hour "mail delivery engines". We need a replacement for SpamCop which sues at least one spammer per month.
The "wearable computing" thing always seems to scream "loser" to me. The first time I encountered true wearable computing was at Alamo rental car turn-in at LAX. Those poor guys wear a computer with an RF link, a credit card reader, a barcode scanner, and a printer. This allows them to check in car after car without a break.
This has been in place for a decade. You don't want that job.
There's been some interest in this in the CAD world. Maintenance techs could see drawings superimposed on their view of the real world.
Stuff like that will probably happen. ("Where does this pipe go? OK, there.") You don't want that job either, but if you have to do it, the gear might help.
Printemps, the Paris department store, had "webcamers" for a few years. Webcamers were cool, young Parisians on Rollerblades wearing designer outfits equipped with audio and video links. Shoppers could make an appointment for remote shopping via the Web. Printemps also sent them to events like the Cannes Film Festival. It was cool, but unprofitable, and has been discontinued.
I've heard Wynn speak, but he didn't have much to say. We got to see what he sees on a big screen.
It wasn't impressive.
We do need better design of wearable electronics. By now there should be phone sets built into earrings, a necklace, and a bracelet.
Yeah, the whole GeForce/Quadro thing is getting tired. They're basically the same chip.
Enforcing the distinction is the only reason for proprietary NVidia drivers. Some features are crippled in the driver when the common driver detects a GeForce card. This is probably the real reasons for the binary-only Linux driver. It also means you can't run many less common OSs on machines with NVidia's NForce chipset, because NVidia uses a common driver for all their hardware.
The most annoying broken feature in the GeForce line is that multiwindow handling is done badly. In Quadro mode, eight overlapping windows are supported in hardware. In GeForce mode, only one is supported. Try running a few OpenGL apps at the same time to see the difference.
It's surprising that NVidia still bothers with the distinction. At one point NVidia bought an interest in ELSA, which was the only remaining seller of Quadro cards. ELSA went bust about two years ago. (I have an ELSA board, with a worthless 7 year warranty.) So there's no high-end wholesale customer to protect.
Now PNY makes Quadro cards, but they're basically an assembly house, not a graphics company, and
the Quadro is a minor part of their business.
If NVidia wanted to have a useful distinction between models, putting more memory on the pro boards would be worthwhile. Animators can use a gigabyte or two of texture memory, because their polygon counts aren't reduced like those of games.
Even if you're doing game work, polygon reduction comes late in the process.
My point is that SSTO designs are so marginal as to be useless. A no-payload SSTO is possible, but not too useful.
If you drop the engines, that counts as a "stage", so it's not single stage to orbit.
Remember Rotary Rocket. They had a plausible SSTO design, and it was about 97% fuel. They had a little weight growth in the engine. That killed it. It couldn't possibly make orbit. Rotary Rocket went bust.
Science has some of the same problems as literary criticism.
Go back to the Kuhn vs. Popper debate to understand this. The hard-line position is that science proceeds by someone proposing a hypothesis, testing it experimentally, encouraging others to test it, and if it survives testing, it moves up to a theory. Hypotheses which are not experimentally falsifiable are not useful. The "soft" position is that science is a cultural construct and hypotheses need not be testable.
Kuhn, the proponent of the "soft" position, won a famous debate on this subject in 1965.
The "soft" position is unpopular because it leads to the conclusion that many "sciences" aren't. Psychology, sociology, and most of economics lose out. So do the "retrospective" fields, like paleontology. They're considered belief systems, not sciences. Since there are more people in those fields than in the hard sciences, this is an unpopular position.
Engineering makes it clear which position is right.
Engineering is based entirely on results which are experimentally falsifiable. Only results tested by experiments which could fail, but didn't, have predictive power. Engineering is about prediction. Without prediction there is no reliability.
Watching C programmers create home-brew objects is pathetic. If you need to encapsulate data structures, use C++.
Yes, C++ has a host of problems and Strostrup and the C++ committee refuse to fix them. But the STL is a huge improvement on malloc/free. (They still can't get auto_ptr right, though.)
"We burned 80 billion barrels of oil this year but discovered less than 10 billion." When we run out of oil, we'll find an alternative. We, as humans, are very clever and creative. It may be a (very) troublesome and painful transition, but we'll make it.
It's been fifty years since the last new major energy source. The "painful transition" may include some big wars.
It's a climate change. There are winners and losers. And the US is a winner.
The Northwest Passage (Atlantic to Pacific, north of Canada) is becoming passable. Icebreakers have forced it in the past, but it's starting to show potential as a shipping route.
The US upper tier of states, and the lower parts of Canada, become more desirable real estate.
More sun, less snow can't hurt.
The US doesn't have that much lowland. The Mississippi Delta and south Florida may be flooded seasonally and during storms, but that happens already. There's going to be some bitching from beachfront property owners, but this is a slow process, slower than a mortgage.
Going to the moon in ten years is a pork program. It only took seven years the first time, after all. It's just a way for Bush to pay off his Texas buddies. Like Reagan and the National Aero$pace Plane.
Space travel with chemical propulsion is never going to get any better. Chemical fuels are as good as they're going to get. There's been essentially zero progress in thirty years.
Building more chemically-fueled spacecraft is a dead end. The weight reduction required for them to work at all makes them so fragile that they'll never be reliable. If you could build a spacecraft with the weight budget of an airliner,
(40% or so of the gross takeoff weight is fuel) spacecraft would be affordable and reliable. But when you have to build something that's 90+% fuel, (SSTO machines are something like 97%+ fuel, which is why nobody has built one), it has to be a fragile balloon full of fuel.
Nuclear power, maybe. But chemical fuels? Been there, done that.
An unmanned lunar orbiter would be worth doing. Last time, in the early 1960s, the US sent five orbiters, which used 70mm film, a chemical film processor, and a scanner to transmit the images back. So they only took 1654 images, and the imagery is only 60 meters per pixel.
Putting a modern survellance camera in lunar orbit would get us 1m imagery of the whole moon, if not better. Maybe we'll find something worth checking out.
I predicted the dot-com collapse with Downside's Deathwatch. Most of the companies listed have went out of business long ago. The remaining ones are mostly just hanging on, with stock prices in the penny-stock range. They're examples of the living dead. Internet America (GEEK) is still around, with the stock around $1 and 82,000 customers left in Texas and Louisiana. Claimsnet.com (CLAI) is at $0.38. There are others. Even after downsizing, the bad-idea companies tend to remain unprofitable. But some of them raised enough money to continue losing at a very modest level for years.
Silicon Valley has many non-public companies that are quietly dying. Often it's not their fault; they were support companies for the semiconductor industry, which has moved elsewhere.
I've been reading "The End of Detroit", on how the US auto industry blew their market share. I see many parallels to Silicon Valley. Auto manufacturing hasn't been centered in Detroit for years now. Detroit, as a city, is a ghost town. The population is half of what it was at peak.
See The Fabulous Ruins of Detroit. That could happen here.
There's already "Patriot Act Compliance Software" for banks. Here's GuardDog from Attus Technologies ("Insuring a Vigilant America").
Automate your compliance procedures for Section 314 (a) of the USA PATRIOT Act and enhance
your anti-money laundering and Bank Security Act procedures. Let GuardDOG, the only software
of its kind, do the tedious work of checking your database on a regular basis against FinCEN's list of
suspicious individuals and businesses.
No Programming Required
User-friendly and easily installed, GuardDOG is a popular FinCEN solution for financial institutions.
The software was designed to identify possible matches with the Financial Crimes Enforcement
Network's (FinCEN) list of suspicious persons, businesses, terrorists and other criminals, while
relieving your IT department of developing a special program compatible with the FinCEN list.
Simply load the CD and you're ready to compare your customer database to the most recent FinCEN
list.
Then there's this scary sounding product:
WatchDOG Photo ID Verification The USA PATRIOT Act mandates that you verify government-issued photo identity for each person you do business with. With our easy-to-use software called WatchDOG Photo ID, financial institutions can now view the most current versions of government-issued photo ID's such as driver's licenses from all U.S. states, U.S. Territories, and I.N.S. travel documents, such as green cards.. In addition to being able to view an exact likeness of each document, a detailed written description is also provided that gives other features of the document that your staff should look for in verifying the document's validity. The software also verifies that the accountholder's driver's license number, for participating states, is a valid number.
As with all ATTUS software, WatchDOG Photo ID is simple to install and can work from any desktop computer.
That last, though, is just a dumb program that shows the general format of photo ID cards. It's just a computerized version of those books club security goons have. It sounds like they are offering access to some nationwide database of ID photos. But they're not. Yet.
The basic problem with hyperthreading is, of course, memory bandwidth. CPUs today are memory-bandwidth starved. 30 years ago, CPUs got about one memory cycle per instruction cycle. Since then, CPUs have speeded up by a factor of about 1000, but memory has only speeded up by a factor of 30 or so. The difference has been papered over, very successfully, with cache.
The cache designers have accomplished more than seems possible. Compare paging to disk, which is a form of cacheing that hasn't improved much in decades.
If you want to benchmark a hyper-threaded machine, a useful exercise is to run two different benchmarks simultaneously. Running the same one is the best case for cache performance; one copy of the benchmark in cache is serving both execution engines. Running different ones lets you see if cache thrashing is occuring. Or
try something like compressing two different video files simultaneously.
If you're seeing significant performance with real-world applications using a a "hyper-threaded" CPU, that's a sign that the operating system's dispatcher is broken. And, of course, hyper-threading dumps more work on the scheduler. There's more stuff to worry about in CPU dispatching now.
Intel seems to be desperate for a new technology that will make people buy new CPUs. The Inanium bombed. The Pentium 4 clock speed hack (faster clock, less performance per clock) has gone as far as it can go. The Pentium 5 seems to be on hold.
Intel doesn't still have a good response to AMD's 64-bit CPUs.
Remember what happened with the Itanium, Intel's last architectural innovation. Intel's plan was to convert the industry over to a technology that couldn't be cloned. This would allow Intel to push CPU price margins back up to their pre-AMD levels. For a few years, Intel had been able to push the price of CPU chips to nearly $1000, and achieved huge margins and profits.
Then came the clones.
Intel has many patents on the innovative technologies of the Itanium. Itanium architecture is different, all right, but not, it's clear by now, better. It's certainly far worse in price/performance. Hyperthreading isn't quite that bad an idea, but it's up there.
From a consumer perspective, it's like four-valve per cylinder auto engines. The performance increase is marginal and it adds some headaches, but it's cool.
Exactly right. Go read "Tog on Interface". The wrong stuff is taking up most of the real estate.
Not only is it yet another combo box, it's a lousy combo box.
The whole "Locations" column is redundant once you've started selecting a file.
Let's see some long filenames and long pathnames in those mockups, and see how it holds up.
The example has no scrollbars. That's unrealistic.
Why do we have "full screen", "minimize", and "close" options on a dialog box? Note that the "Cancel" and "close" buttons typically do the same thing.
Slashdot Mirror
That's a good spam-filtering algorithm, too. As I keep telling people who fight spam, "follow the money". Quit worrying about where the spam is coming from. Follow where the money goes.
The defining moment was when they pulled the plug in Berkeley's BSD group in the early 1980s (DARPA decided to fund Mach instead) and BSD went on anyway, with private funding.
In retrospect, Xerox had it right in XNS - 48-bit MAP addresses on the LAN, and 48-bit net numbers for routing between LANs. When the transition to IP came along, the old ARPANET lobby wanted to just transition by putting their IMP number in the second half of the IP address, and adding [10.0.xxx.xxx]. That's how we got into this mess of class A, B and C networks, netblocks, NAT, and all this other junk.
IPv6 is in some ways worse, because the interpretation of those 128 bits is complicated. Not everybody gets an autonomous system number and gets to participate in routing.
Bitpass is either a depository institution, like a bank, or a "check seller", like companies that sell money orders. Either way, they should have the appropriate license and be subject to bank regulations and audits. You can't just go into business handling other people's money without oversight.
I want to subscribe to a service which sues spammers. The CAN-SPAM act's definition of an ISP seems to include a service like SpamCop. But SpamCop doesn't have a litigation staff, and their parent, IronPort sells spammer-friendly million-email-per-hour "mail delivery engines". We need a replacement for SpamCop which sues at least one spammer per month.
There's been some interest in this in the CAD world. Maintenance techs could see drawings superimposed on their view of the real world. Stuff like that will probably happen. ("Where does this pipe go? OK, there.") You don't want that job either, but if you have to do it, the gear might help.
Printemps, the Paris department store, had "webcamers" for a few years. Webcamers were cool, young Parisians on Rollerblades wearing designer outfits equipped with audio and video links. Shoppers could make an appointment for remote shopping via the Web. Printemps also sent them to events like the Cannes Film Festival. It was cool, but unprofitable, and has been discontinued.
I've heard Wynn speak, but he didn't have much to say. We got to see what he sees on a big screen. It wasn't impressive.
We do need better design of wearable electronics. By now there should be phone sets built into earrings, a necklace, and a bracelet.
Enforcing the distinction is the only reason for proprietary NVidia drivers. Some features are crippled in the driver when the common driver detects a GeForce card. This is probably the real reasons for the binary-only Linux driver. It also means you can't run many less common OSs on machines with NVidia's NForce chipset, because NVidia uses a common driver for all their hardware.
The most annoying broken feature in the GeForce line is that multiwindow handling is done badly. In Quadro mode, eight overlapping windows are supported in hardware. In GeForce mode, only one is supported. Try running a few OpenGL apps at the same time to see the difference.
It's surprising that NVidia still bothers with the distinction. At one point NVidia bought an interest in ELSA, which was the only remaining seller of Quadro cards. ELSA went bust about two years ago. (I have an ELSA board, with a worthless 7 year warranty.) So there's no high-end wholesale customer to protect. Now PNY makes Quadro cards, but they're basically an assembly house, not a graphics company, and the Quadro is a minor part of their business.
If NVidia wanted to have a useful distinction between models, putting more memory on the pro boards would be worthwhile. Animators can use a gigabyte or two of texture memory, because their polygon counts aren't reduced like those of games. Even if you're doing game work, polygon reduction comes late in the process.
If you want to prototype, there's Mecchano/Erector, and there are some higher-quality kids from Berg and 80/20.
Why am I in Wally Cleaver's bedroom?
Wasn't this all covered in yesterday's SCO story?
If you want to make stuff, get a milling machine like everybody else. Legos are for kids.
The people who wear that stuff usually look like losers.
Remember Rotary Rocket. They had a plausible SSTO design, and it was about 97% fuel. They had a little weight growth in the engine. That killed it. It couldn't possibly make orbit. Rotary Rocket went bust.
The "soft" position is unpopular because it leads to the conclusion that many "sciences" aren't. Psychology, sociology, and most of economics lose out. So do the "retrospective" fields, like paleontology. They're considered belief systems, not sciences. Since there are more people in those fields than in the hard sciences, this is an unpopular position.
Engineering makes it clear which position is right. Engineering is based entirely on results which are experimentally falsifiable. Only results tested by experiments which could fail, but didn't, have predictive power. Engineering is about prediction. Without prediction there is no reliability.
True. There are 1.8 million Clementine images. More resolution might be nice, but the existing imagery hasn't been studied fully yet.
Yes, C++ has a host of problems and Strostrup and the C++ committee refuse to fix them. But the STL is a huge improvement on malloc/free. (They still can't get auto_ptr right, though.)
It's been fifty years since the last new major energy source. The "painful transition" may include some big wars.
Space travel with chemical propulsion is never going to get any better. Chemical fuels are as good as they're going to get. There's been essentially zero progress in thirty years.
Building more chemically-fueled spacecraft is a dead end. The weight reduction required for them to work at all makes them so fragile that they'll never be reliable. If you could build a spacecraft with the weight budget of an airliner, (40% or so of the gross takeoff weight is fuel) spacecraft would be affordable and reliable. But when you have to build something that's 90+% fuel, (SSTO machines are something like 97%+ fuel, which is why nobody has built one), it has to be a fragile balloon full of fuel.
Nuclear power, maybe. But chemical fuels? Been there, done that.
An unmanned lunar orbiter would be worth doing. Last time, in the early 1960s, the US sent five orbiters, which used 70mm film, a chemical film processor, and a scanner to transmit the images back. So they only took 1654 images, and the imagery is only 60 meters per pixel. Putting a modern survellance camera in lunar orbit would get us 1m imagery of the whole moon, if not better. Maybe we'll find something worth checking out.
Silicon Valley has many non-public companies that are quietly dying. Often it's not their fault; they were support companies for the semiconductor industry, which has moved elsewhere.
I've been reading "The End of Detroit", on how the US auto industry blew their market share. I see many parallels to Silicon Valley. Auto manufacturing hasn't been centered in Detroit for years now. Detroit, as a city, is a ghost town. The population is half of what it was at peak. See The Fabulous Ruins of Detroit. That could happen here.
Then there's this scary sounding product:
-
WatchDOG Photo ID Verification
That last, though, is just a dumb program that shows the general format of photo ID cards. It's just a computerized version of those books club security goons have. It sounds like they are offering access to some nationwide database of ID photos. But they're not. Yet.The USA PATRIOT Act mandates that you verify government-issued photo identity for each person you do business with. With our easy-to-use software called WatchDOG Photo ID, financial institutions can now view the most current versions of government-issued photo ID's such as driver's licenses from all U.S. states, U.S. Territories, and I.N.S. travel documents, such as green cards.. In addition to being able to view an exact likeness of each document, a detailed written description is also provided that gives other features of the document that your staff should look for in verifying the document's validity. The software also verifies that the accountholder's driver's license number, for participating states, is a valid number.
As with all ATTUS software, WatchDOG Photo ID is simple to install and can work from any desktop computer.
That's how we got into this mess. Nader took votes away from Gore. Remember?
Intel originally announced the 386 as "intended for servers". They did the same thing with the Pentium Pro, their first real superscalar.
For a while, Dell and HP sold Itanium desktops, but nobody bought.
If you want to benchmark a hyper-threaded machine, a useful exercise is to run two different benchmarks simultaneously. Running the same one is the best case for cache performance; one copy of the benchmark in cache is serving both execution engines. Running different ones lets you see if cache thrashing is occuring. Or try something like compressing two different video files simultaneously.
If you're seeing significant performance with real-world applications using a a "hyper-threaded" CPU, that's a sign that the operating system's dispatcher is broken. And, of course, hyper-threading dumps more work on the scheduler. There's more stuff to worry about in CPU dispatching now.
Intel seems to be desperate for a new technology that will make people buy new CPUs. The Inanium bombed. The Pentium 4 clock speed hack (faster clock, less performance per clock) has gone as far as it can go. The Pentium 5 seems to be on hold. Intel doesn't still have a good response to AMD's 64-bit CPUs.
Remember what happened with the Itanium, Intel's last architectural innovation. Intel's plan was to convert the industry over to a technology that couldn't be cloned. This would allow Intel to push CPU price margins back up to their pre-AMD levels. For a few years, Intel had been able to push the price of CPU chips to nearly $1000, and achieved huge margins and profits. Then came the clones.
Intel has many patents on the innovative technologies of the Itanium. Itanium architecture is different, all right, but not, it's clear by now, better. It's certainly far worse in price/performance. Hyperthreading isn't quite that bad an idea, but it's up there.
From a consumer perspective, it's like four-valve per cylinder auto engines. The performance increase is marginal and it adds some headaches, but it's cool.
The whole "Locations" column is redundant once you've started selecting a file.
Let's see some long filenames and long pathnames in those mockups, and see how it holds up. The example has no scrollbars. That's unrealistic.
Why do we have "full screen", "minimize", and "close" options on a dialog box? Note that the "Cancel" and "close" buttons typically do the same thing.