Although 9600 could never handle today's internet and web activities, it is amazingly fast for TTY and CLI type applications. Having started with 110 baud mechanical TTY and 300 baud acoustical coupler modem on a green screen, I well remember my first experience with a 9600 baud hardwired Lear Siegler terminals -- WOW very fast.
9600 baud is good enough for modem-to-modem chat, e-mail via pine, text processing with vi or emacs, or almost any *nix command. Thinking about this reminds me of how terribly bloated everything has become with verbose formatting and styling of pages. Pictures may be worth a 1000 words, but they require 10 to 100 times the bandwidth of those words.
If more people published more good ideas in the public domain, businesses would have less room for silly software patents. This publication process would need to work with, educate, and support patent examiners -- making it easier for them to deny the more egregious claims before they are issued. And if thousands OSS fanatics can't come up with the idea to keep it out of the clutches of patent-happy companies, then perhaps it was sufficiently innovative and original that it merits a financial rewards of a patent.
I use one 250 GB firewire drive for "dailies" using rsync to backup changed files. About once a month or so, I wipe that drive clean and use Carbon Copy Cloner to make a full bootable backup.
I also bought 2 bare drives ( 250GB Maxtor MaXLine Plus II 7200rpm 9ms 8MB Cache Parallel ATA ATA/133), a Miglia Catalyst ATA/FW Conversion Kit (its a little circuit board that mounts inside an old external SCSI drive enclosure and converts it to use any ATA drive on firewire), and an ancient external SCSI drive with a slip-off cover. I love using the easy-to-open external firewire enclosure. Not only is it great for backups, but its nice for doing large file transfers and troubleshooting between computers (just remove a drive from a computer, pop it in the enclosure and mount it on any firewire-capable machine).
For backups, I use Carbon Copy Cloner to make a full bootable backup of my drive onto one of the bare drives sitting inside the enclosure. Doing a full backup of 140 GB takes about 2 hours (I run it overnight) with about 10-15 minutes labor on either end to setup and verify the thing (I both check the filesystem on the backup and test boot it). I like having a bootable backup because in an emergency I'd hate to have to reinstall/reconfig the OS.
After I make a full backup, I remove the bare drive from enclosure and take it (in an antistatic bubblewrap bag) to the bank safe deposit box and bring back the bare drive that was in off-site storage. Bare drives fit nicely in a safe deposit box. I prefer the ping-pong method for offsites because it means that there is always a safe, offsite copy of my files.
If I were starting over, I'd probably go with 3 bare drives (keeping 2 offsite at all times) and an enclosure. If I weren't paranoid about hardware failures I'd go with 2 drives and an enclosure, but I like the idea that if the hardware fried during a backup (killing both the internal HD and the in-process backup) that I'd have both a safe offsite and onsite backup.
I've always thought that there might be an intentional diversity in the genetic components of human behavior, not unlike the hypervariablity found in the genetics underlying the immune system. Human society functions better in a nonstationary environment (= ice ages, floods, dry spells, changes in diet from whale blubber to potatoes) if the society is structurally non-homogeneous. Society needs risk takers and risk avoiders, optimists and pessimists, manic spenders and thrifty savers, lone achievers and gregarious team players. How else can we cope with the rich times, the poor times, the peace times, war times, the stay-at-home times, and the move-to-another-land-times.
Anytime I find a useful web page, I cache a copy on my local disk (strictly for personal/ time shifted use only). Personal sites have a way of disappearing when the owner loses interest and corporate sites have a tendency to flush old data when they reorg a site. Even with the cost of backups (I use 3 x 250 GB HDs for onsite/offsite backup), keeping a local copy costs a few tenths of a penny per MB.
The loss of old content is sad, really. The web is sometimes more like the spoken word than the written word. It is ephemeral -- if you weren't there when the page was posted, you have a high chance of never getting it.
There are already lots of patents on quantum computing:
I actually like these patents because they will die before quantum computing gets off the ground (that first patent you cite is from 1993). They may cause some frustration for researchers and pioneers, but they won't encumber widespread commercialization 20 years from now.
As an aside, I suspect it will take much more than 10-20 years for quantum computing to overtake semiconductor computers. Old technologies have a way of holding on and remaining competitive by virtue of the install base of engineers, businesses, and users. Only when quantum computing can do everything (and more) than semiconductor computing will people seriously consider it. Even then, there will be a lag as engineers and business learn how to commercialize the product.
I hate to be cynical, but who's funding this kind of research, directly or indirectly? Now think about this strategy:....
That's very true, but what I am talking about are the obvious patents, not the ones that require millions of dollars in legitimate investment in R&D. I'm talking about silly little patents that take someone a few hours of thinking and then they try to claim any use of quantum mechanics in some broad area of endeavor (like using qubits to optimize internet routings, or using entanglement to serve ads, or some such "add-a-q-to-any-ordinary-activity" type of patent).
Personally, I am in favor of patents for non-trivial inventions. I wonder if part of the problem with the current patent system is that the examiners may not understand the state of the art well enough to judge which inventions were obvious and which inventions were hard. The point is that easy inventions don't need the encouragement created by a patent -- they will get invented and deployed anyway. Patents should reserved for inventions that could not have happened if the inventor did not think they had a chance of a patent.
It's a separate issue, entirely, whether the fruits of publicly funded research should be patented at all, but that's a different discussion.
Given the patent fiasco of the internet (just add "e" to anything and receive a free patent), now is the time to create prior art for quantum computing and publish all the ideas for adding "q" to everything. Only by striking first and getting innovation in the public domain can we have true open and unencumbered standards.
And as long as wide spread adoption of quantum computing is more that 17 years away, companies can't read this message and strike first (prepatenting these ideas first). If companies patent ideas too soon, the patent will be dead when the real money is being made.
I've often wondered about the rate of evaporation on fluid-bearing HDs. Running at high temperatures, even the lowest level of evaporation will mean a loss of lubricant over the years. Even inside the semi-sealed chamber of the drive, each on-off cycle will mean that the drive exhales the evaporated lubricant and the cooling pattern of a turned-off drive will mean condensation of lubricant on the inside shell of the case.
Ball-bearings (well-designed ones at least) can last virtually forever. I wonder if the same can be said for fluid bearings?
Although the photos may be fake, the design is sound. A 17 inch or 20 inch LCD creates a massive space of real estate in the back -- plenty of room for a mobo, powersupply, drive bays, etc.
Cooling would only be trick if Apple insists on passive cooling (which is just the type of thing Jobs might do). Even this would not be that hard an engineering problem because the height of the monitor provides an excellent chimney effect for passive cooling. Mounting the G5 (and other hot electronics) at the base of the unit and allowing the G5-heated air to rise the height of the monitor would provide a fairly nice draft (at least 4-8 times the cooling capacity of the ill-fated PowerMac Cube). An auxillary fan might be used if needed (a large low-RPM exhaust fan in the center top could provide extra airflow if needed).
Thanks for the link! And I thank the others that responded to this thread, providing insights about and sightings of these Scrollpoint mice.
I had tried to Google for this type of product, but without knowing the "scrollpoint" brand name, the search was futile. This situation made me realize that there is a fundamental flaw in both search engines and branding --its very hard to search for something that someone else has labeled with an unknown name. Perhaps someone will produce an ontology of brand names (defining each brand name object in terms of other brands names and plain language). Perhaps somebody already has, but I don't know the name of it.;)
IBM has a prototype of a mouse with trackpoint scroll stick. Because the trackpoint nubbin is a rate-device, like a joystick, it apparently offers superior productivity to a scrollwheel according to IBM's research (PDF of slides).
Has anyone seen any devices like this? As much as I love the scrollwheel, my finger gets tired scrolling through a long document -- I'd rather just pull on a stick/nubbin and zoom along.
Many animals display social behavior - from ants and termites to blue jays, llamas, dolphins, monkeys, and people. Is it really so surprising that these organisms (including people) might have a built-in, evolved accounting system for social relationships -- if A cheats me, it feels good the cheat A back. The basic tit-for-tat strategy is very well known in iterated game theory so its no surprise that it might be hardwired into social organisms.
This model is very compelling for the commercial market -- companies know that they will both want customization and will need support for their software. They are willing to pay for expert assistance and 7x24 access to services. Enterprise software and support can sell for hundreds or thousands of dollars per seat - providing plenty of revenues to offset the labor costs of support can customization.
But the consumer market is very different. The consumer market has very low retail prices that can't support the high cost of labor - a $49.95 price point product can go from profit to loss on a single tech support call. This consumer market consists of two segments -- geeks who don't need support and the clueless who needs lots of expensive support. Currently, proprietary software makers can earn a profit, in aggregate, because they capture money from both the geek and clueless segments. They may lose money on the clueless, but that make up for it on the geeks who don't need support.
In a FOSS environment, the geeks can go for the free downloads and do-it-themselves when it comes to deployment, customization, and support of FOSS. Geeks have little reason to pay for FOSS-related services. This leaves only the labor-intensive clueless expecting to get a year of support for their $49.95. But because they are clueless, they will use more that $49.95 of support labor (even if that labor is in India).
The trick with these services models is finding people that are both willing to pay for service but that don't actually need to use the service that much. Its a very good model for corporate IT, but I don't see how the numbers can work on the consumer side. Perhaps someone in tech support has numbers for the statistical distribution of the percentages of people that use X-minutes of support.
If they added GPS to the mix and an autotrack function (with "do not disturb/do not track" toggle, of course) then people could use the service without having to stop all the time and text the server. The minute you move more than 50 feet from your "official" location, the GPS would recompute and resend a new update. As long as you are in motion, it sends a "Not stationary" message. Once you arrive, it notices the stabilization in position and sends the new locale (maybe reverse lookup to provide a street addy or the name of the club).
This same idea can be used make underwater buildings. It only takes about 1 kWh to make 4.2 lbs of seacrete and its about as strong as sidewalk concrete.
As much as I hate cigarette smoke, I'm not sure I want robots running around yanking cigarettes from people's mouths. After all, letting someone smoke would clearly be a violation of the "harm through inaction " law of robotics. Society already mandates the removal of too much personal risk and self-responsibility. The last thing we need is robots deciding what their human "masters" can and cannot do.
With 220 million transistors, do we need Intel?
on
SIGGraph and Open Source
·
· Score: 2, Interesting
I notice that the NVIDIA 6800 has 220 million transistors. If they added a few million more transistors or sacrificed a few pipes for a RISC processor, the chip could do all the computation for the system. For people whose only demanding applications are graphics-intensive games, a CPU-on-GPU design might be a great idea. Admittedly, this solution does nothing for servers, but then it does not seem like servers are driving the mass-market PC technology at the moment.
This is interesting and promising technology. But like all antispam techniques, spammers will find a way around it. Once spammers get a copy of the software, they can create and test countermeasures in the comfort of their own sleazy lairs.
For example, the article mentions the software accepts a message that is long but has a few "spammy" sequences. This suggests an immediate countermeasure of adding bulk to spam -- appending a copy of some news article to the spammy payload (some already do this).
Personally, I've always thought that a simple spell check would do a good job as another layer filtering. It would place spammers in a no-win situation -- either the keyword filter or the spell check filter would get them.
I wonder if it is legally possible to mount a class-action defense? The defendants could then pool their resources for lawyerage, expert witnesses, etc. If a class of parties can act as a plaintiff, why can't a similar structure be used in defense?
I know it's a bit of an anti-establishment thought, but I'm not sure demand-side pricing is ethical. The whole idea of trying to take your customers for everything you can sounds so much colder when you look at it from their side.
And the whole idea of trying to take your suppliers for everything by demanding the lowest possible price is pretty cold (just ask the suppliers to Wal-Mart or Dell).
Neither supply-side nor demand-side pricing is wholly ethical - it depends on your perspective. If I find someone selling something for $1 and I know I can do something with that item that brings me $100 (or saves a $100 in costs), shouldn't I share the wealth with my supplier. In that scenario the customer is unethical in trying to get the lowest cost possible (taking the supplier for all they can). Just because it cost someone $1 (supply-side price) does not mean its worth $1 of the fair share of the value it creates.
You mention the notion of a fair markup - it would seem that both the supplier and the customer should agree on a mutual mark-up so that the profit to the supplier and the profit that the supplier's product brings to the customer are "fair." That would be the ethical thing to do.
LOAF sounds wonderful until someone creates a LOAF-exploiting virus. If a friend becomes infected, their 0wned machine can send virus messages (with the friend's LOAF signature) that have a very high chance of being read and thus spreading through a LOAF network.
The challenge with any computer-based social network is not the "do I trust my friend" question but the issue of "do I trust my friend's computer that is sending me this message"? Perhaps all computers need a tamperproof hash that encodes their OS patch/AV update/spyware/firewall defense state. That way the message recipient can assess the trustworthyness of the sending machine.
Does the Internet Need a DoD/police Force?
on
The Spyware Inferno
·
· Score: 3, Interesting
National governments seem clueless/powerless/apathetic with regard to malware (spyware, phishing, viruses, etc.) The current ad hoc approach - independent semi-commercialized tracking/alert/filtering services don't do a very good job, provide less than 100% coverage (of both PCs and treats), suffer from lack due process (e.g., how does a nonspammer get unblacklisted), and are purely passive (doing nothing to halt spammers, phishers, etc.)
I wonder when the users of the internet will form their own supranational government, with a defense force and coordinated policing actvities. Taxation might be in the form of CPU cycles & bandwidth used by policing actions to DDoS convicted spammers/phishers/spyware providers.
So many of the arguments against AI take the form "what if something unexpected happens" and the AI is too dumb to do the right thing. Its a valid issue but one that has a simple counter-argument in the case of space exploration. What if something unexpected happens and the rover does not have a hour to wait for an intelligent answer? Sometimes a late decision is as bad as the wrong decision.
Moreover, in the context of space probes, long distance bandwidth limitations means that the local AI has far more data at far faster response times than do the mission controllers. While the mission controllers wait to download 3 carefully chosen snapshots of the terrain or obstacle, the local AI could be interpreting 30 fps visual data as it moves. The local AI may be dumb, but bandwidth limits make the mission controller dumber.
As for the "unproven" problem, this can be remedied by building autonomous Earth rovers and letting them run around and "discover" the Earth. They might even make nice deep sea exploration vehicles.
Slashdot Mirror
Although 9600 could never handle today's internet and web activities, it is amazingly fast for TTY and CLI type applications. Having started with 110 baud mechanical TTY and 300 baud acoustical coupler modem on a green screen, I well remember my first experience with a 9600 baud hardwired Lear Siegler terminals -- WOW very fast.
9600 baud is good enough for modem-to-modem chat, e-mail via pine, text processing with vi or emacs, or almost any *nix command. Thinking about this reminds me of how terribly bloated everything has become with verbose formatting and styling of pages. Pictures may be worth a 1000 words, but they require 10 to 100 times the bandwidth of those words.
If more people published more good ideas in the public domain, businesses would have less room for silly software patents. This publication process would need to work with, educate, and support patent examiners -- making it easier for them to deny the more egregious claims before they are issued. And if thousands OSS fanatics can't come up with the idea to keep it out of the clutches of patent-happy companies, then perhaps it was sufficiently innovative and original that it merits a financial rewards of a patent.
I use one 250 GB firewire drive for "dailies" using rsync to backup changed files. About once a month or so, I wipe that drive clean and use Carbon Copy Cloner to make a full bootable backup.
I also bought 2 bare drives ( 250GB Maxtor MaXLine Plus II 7200rpm 9ms 8MB Cache Parallel ATA ATA/133), a Miglia Catalyst ATA/FW Conversion Kit (its a little circuit board that mounts inside an old external SCSI drive enclosure and converts it to use any ATA drive on firewire), and an ancient external SCSI drive with a slip-off cover. I love using the easy-to-open external firewire enclosure. Not only is it great for backups, but its nice for doing large file transfers and troubleshooting between computers (just remove a drive from a computer, pop it in the enclosure and mount it on any firewire-capable machine).
For backups, I use Carbon Copy Cloner to make a full bootable backup of my drive onto one of the bare drives sitting inside the enclosure. Doing a full backup of 140 GB takes about 2 hours (I run it overnight) with about 10-15 minutes labor on either end to setup and verify the thing (I both check the filesystem on the backup and test boot it). I like having a bootable backup because in an emergency I'd hate to have to reinstall/reconfig the OS.
After I make a full backup, I remove the bare drive from enclosure and take it (in an antistatic bubblewrap bag) to the bank safe deposit box and bring back the bare drive that was in off-site storage. Bare drives fit nicely in a safe deposit box. I prefer the ping-pong method for offsites because it means that there is always a safe, offsite copy of my files.
If I were starting over, I'd probably go with 3 bare drives (keeping 2 offsite at all times) and an enclosure. If I weren't paranoid about hardware failures I'd go with 2 drives and an enclosure, but I like the idea that if the hardware fried during a backup (killing both the internal HD and the in-process backup) that I'd have both a safe offsite and onsite backup.
I've always thought that there might be an intentional diversity in the genetic components of human behavior, not unlike the hypervariablity found in the genetics underlying the immune system. Human society functions better in a nonstationary environment (= ice ages, floods, dry spells, changes in diet from whale blubber to potatoes) if the society is structurally non-homogeneous. Society needs risk takers and risk avoiders, optimists and pessimists, manic spenders and thrifty savers, lone achievers and gregarious team players. How else can we cope with the rich times, the poor times, the peace times, war times, the stay-at-home times, and the move-to-another-land-times.
Anytime I find a useful web page, I cache a copy on my local disk (strictly for personal/ time shifted use only). Personal sites have a way of disappearing when the owner loses interest and corporate sites have a tendency to flush old data when they reorg a site. Even with the cost of backups (I use 3 x 250 GB HDs for onsite/offsite backup), keeping a local copy costs a few tenths of a penny per MB.
The loss of old content is sad, really. The web is sometimes more like the spoken word than the written word. It is ephemeral -- if you weren't there when the page was posted, you have a high chance of never getting it.
There are already lots of patents on quantum computing:
I actually like these patents because they will die before quantum computing gets off the ground (that first patent you cite is from 1993). They may cause some frustration for researchers and pioneers, but they won't encumber widespread commercialization 20 years from now.
As an aside, I suspect it will take much more than 10-20 years for quantum computing to overtake semiconductor computers. Old technologies have a way of holding on and remaining competitive by virtue of the install base of engineers, businesses, and users. Only when quantum computing can do everything (and more) than semiconductor computing will people seriously consider it. Even then, there will be a lag as engineers and business learn how to commercialize the product.
I hate to be cynical, but who's funding this kind of research, directly or indirectly? Now think about this strategy:....
That's very true, but what I am talking about are the obvious patents, not the ones that require millions of dollars in legitimate investment in R&D. I'm talking about silly little patents that take someone a few hours of thinking and then they try to claim any use of quantum mechanics in some broad area of endeavor (like using qubits to optimize internet routings, or using entanglement to serve ads, or some such "add-a-q-to-any-ordinary-activity" type of patent).
Personally, I am in favor of patents for non-trivial inventions. I wonder if part of the problem with the current patent system is that the examiners may not understand the state of the art well enough to judge which inventions were obvious and which inventions were hard. The point is that easy inventions don't need the encouragement created by a patent -- they will get invented and deployed anyway. Patents should reserved for inventions that could not have happened if the inventor did not think they had a chance of a patent.
It's a separate issue, entirely, whether the fruits of publicly funded research should be patented at all, but that's a different discussion.
Given the patent fiasco of the internet (just add "e" to anything and receive a free patent), now is the time to create prior art for quantum computing and publish all the ideas for adding "q" to everything. Only by striking first and getting innovation in the public domain can we have true open and unencumbered standards.
And as long as wide spread adoption of quantum computing is more that 17 years away, companies can't read this message and strike first (prepatenting these ideas first). If companies patent ideas too soon, the patent will be dead when the real money is being made.
I've often wondered about the rate of evaporation on fluid-bearing HDs. Running at high temperatures, even the lowest level of evaporation will mean a loss of lubricant over the years. Even inside the semi-sealed chamber of the drive, each on-off cycle will mean that the drive exhales the evaporated lubricant and the cooling pattern of a turned-off drive will mean condensation of lubricant on the inside shell of the case.
Ball-bearings (well-designed ones at least) can last virtually forever. I wonder if the same can be said for fluid bearings?
Although the photos may be fake, the design is sound. A 17 inch or 20 inch LCD creates a massive space of real estate in the back -- plenty of room for a mobo, powersupply, drive bays, etc.
Cooling would only be trick if Apple insists on passive cooling (which is just the type of thing Jobs might do). Even this would not be that hard an engineering problem because the height of the monitor provides an excellent chimney effect for passive cooling. Mounting the G5 (and other hot electronics) at the base of the unit and allowing the G5-heated air to rise the height of the monitor would provide a fairly nice draft (at least 4-8 times the cooling capacity of the ill-fated PowerMac Cube). An auxillary fan might be used if needed (a large low-RPM exhaust fan in the center top could provide extra airflow if needed).
Thanks for the link! And I thank the others that responded to this thread, providing insights about and sightings of these Scrollpoint mice.
;)
I had tried to Google for this type of product, but without knowing the "scrollpoint" brand name, the search was futile. This situation made me realize that there is a fundamental flaw in both search engines and branding --its very hard to search for something that someone else has labeled with an unknown name. Perhaps someone will produce an ontology of brand names (defining each brand name object in terms of other brands names and plain language). Perhaps somebody already has, but I don't know the name of it.
IBM has a prototype of a mouse with trackpoint scroll stick. Because the trackpoint nubbin is a rate-device, like a joystick, it apparently offers superior productivity to a scrollwheel according to IBM's research (PDF of slides).
Has anyone seen any devices like this? As much as I love the scrollwheel, my finger gets tired scrolling through a long document -- I'd rather just pull on a stick/nubbin and zoom along.
Many animals display social behavior - from ants and termites to blue jays, llamas, dolphins, monkeys, and people. Is it really so surprising that these organisms (including people) might have a built-in, evolved accounting system for social relationships -- if A cheats me, it feels good the cheat A back. The basic tit-for-tat strategy is very well known in iterated game theory so its no surprise that it might be hardwired into social organisms.
This model is very compelling for the commercial market -- companies know that they will both want customization and will need support for their software. They are willing to pay for expert assistance and 7x24 access to services. Enterprise software and support can sell for hundreds or thousands of dollars per seat - providing plenty of revenues to offset the labor costs of support can customization.
But the consumer market is very different. The consumer market has very low retail prices that can't support the high cost of labor - a $49.95 price point product can go from profit to loss on a single tech support call. This consumer market consists of two segments -- geeks who don't need support and the clueless who needs lots of expensive support. Currently, proprietary software makers can earn a profit, in aggregate, because they capture money from both the geek and clueless segments. They may lose money on the clueless, but that make up for it on the geeks who don't need support.
In a FOSS environment, the geeks can go for the free downloads and do-it-themselves when it comes to deployment, customization, and support of FOSS. Geeks have little reason to pay for FOSS-related services. This leaves only the labor-intensive clueless expecting to get a year of support for their $49.95. But because they are clueless, they will use more that $49.95 of support labor (even if that labor is in India).
The trick with these services models is finding people that are both willing to pay for service but that don't actually need to use the service that much. Its a very good model for corporate IT, but I don't see how the numbers can work on the consumer side. Perhaps someone in tech support has numbers for the statistical distribution of the percentages of people that use X-minutes of support.
If they added GPS to the mix and an autotrack function (with "do not disturb/do not track" toggle, of course) then people could use the service without having to stop all the time and text the server. The minute you move more than 50 feet from your "official" location, the GPS would recompute and resend a new update. As long as you are in motion, it sends a "Not stationary" message. Once you arrive, it notices the stabilization in position and sends the new locale (maybe reverse lookup to provide a street addy or the name of the club).
Just don't tell your employer that you have this.
Nothing for you to see here. Please move along.
/. home page. I guess I need that worm to read slashdot now.
That's what I saw when I first clicked on the link to "Read More" for the
This same idea can be used make underwater buildings. It only takes about 1 kWh to make 4.2 lbs of seacrete and its about as strong as sidewalk concrete.
Its time to grow an underwater home!
As much as I hate cigarette smoke, I'm not sure I want robots running around yanking cigarettes from people's mouths. After all, letting someone smoke would clearly be a violation of the "harm through inaction " law of robotics. Society already mandates the removal of too much personal risk and self-responsibility. The last thing we need is robots deciding what their human "masters" can and cannot do.
I notice that the NVIDIA 6800 has 220 million transistors. If they added a few million more transistors or sacrificed a few pipes for a RISC processor, the chip could do all the computation for the system. For people whose only demanding applications are graphics-intensive games, a CPU-on-GPU design might be a great idea. Admittedly, this solution does nothing for servers, but then it does not seem like servers are driving the mass-market PC technology at the moment.
This is interesting and promising technology. But like all antispam techniques, spammers will find a way around it. Once spammers get a copy of the software, they can create and test countermeasures in the comfort of their own sleazy lairs.
For example, the article mentions the software accepts a message that is long but has a few "spammy" sequences. This suggests an immediate countermeasure of adding bulk to spam -- appending a copy of some news article to the spammy payload (some already do this).
Personally, I've always thought that a simple spell check would do a good job as another layer filtering. It would place spammers in a no-win situation -- either the keyword filter or the spell check filter would get them.
I wonder if it is legally possible to mount a class-action defense? The defendants could then pool their resources for lawyerage, expert witnesses, etc. If a class of parties can act as a plaintiff, why can't a similar structure be used in defense?
I know it's a bit of an anti-establishment thought, but I'm not sure demand-side pricing is ethical. The whole idea of trying to take your customers for everything you can sounds so much colder when you look at it from their side.
And the whole idea of trying to take your suppliers for everything by demanding the lowest possible price is pretty cold (just ask the suppliers to Wal-Mart or Dell).
Neither supply-side nor demand-side pricing is wholly ethical - it depends on your perspective. If I find someone selling something for $1 and I know I can do something with that item that brings me $100 (or saves a $100 in costs), shouldn't I share the wealth with my supplier. In that scenario the customer is unethical in trying to get the lowest cost possible (taking the supplier for all they can). Just because it cost someone $1 (supply-side price) does not mean its worth $1 of the fair share of the value it creates.
You mention the notion of a fair markup - it would seem that both the supplier and the customer should agree on a mutual mark-up so that the profit to the supplier and the profit that the supplier's product brings to the customer are "fair." That would be the ethical thing to do.
LOAF sounds wonderful until someone creates a LOAF-exploiting virus. If a friend becomes infected, their 0wned machine can send virus messages (with the friend's LOAF signature) that have a very high chance of being read and thus spreading through a LOAF network.
The challenge with any computer-based social network is not the "do I trust my friend" question but the issue of "do I trust my friend's computer that is sending me this message"? Perhaps all computers need a tamperproof hash that encodes their OS patch/AV update/spyware/firewall defense state. That way the message recipient can assess the trustworthyness of the sending machine.
National governments seem clueless/powerless/apathetic with regard to malware (spyware, phishing, viruses, etc.) The current ad hoc approach - independent semi-commercialized tracking/alert/filtering services don't do a very good job, provide less than 100% coverage (of both PCs and treats), suffer from lack due process (e.g., how does a nonspammer get unblacklisted), and are purely passive (doing nothing to halt spammers, phishers, etc.)
I wonder when the users of the internet will form their own supranational government, with a defense force and coordinated policing actvities. Taxation might be in the form of CPU cycles & bandwidth used by policing actions to DDoS convicted spammers/phishers/spyware providers.
So many of the arguments against AI take the form "what if something unexpected happens" and the AI is too dumb to do the right thing. Its a valid issue but one that has a simple counter-argument in the case of space exploration. What if something unexpected happens and the rover does not have a hour to wait for an intelligent answer? Sometimes a late decision is as bad as the wrong decision.
Moreover, in the context of space probes, long distance bandwidth limitations means that the local AI has far more data at far faster response times than do the mission controllers. While the mission controllers wait to download 3 carefully chosen snapshots of the terrain or obstacle, the local AI could be interpreting 30 fps visual data as it moves. The local AI may be dumb, but bandwidth limits make the mission controller dumber.
As for the "unproven" problem, this can be remedied by building autonomous Earth rovers and letting them run around and "discover" the Earth. They might even make nice deep sea exploration vehicles.