Perhaps the problem is that the internet is too massively large for the human mind, social systems, and trademark laws to handle. Everyone thinks they are coming up with a unique, non-overlapping, name from everyone else. But once the system becomes too larger, very few names are unique. In reality, all the "good" names are taken and even all the easy variants of the good names are taken. Its a case of too many people and too few names.
The case of confusing/typoed near-names (ggle.com) is also a human scalability problem. If one only interacted within a tribe or small group (say 100 individuals), a typo or near-name would still be unambiguous.
People (and their social/legal systems) weren't designed to connect directly to millions or billions of others.
These techniques seem like brute force schemes to deal with the thermal resistance of chip packages -- you have to cool the heatsink to -110C in order to keep the "intel inside" at less than +60C). Why not use backside thinning. to bring the hot circuits of the processor within microns of a high coolant flux chamber. Backside thinning could get the coolant to within 10 microns of the junctions. If the CCD people can thin a massive 2k x 2k CCDs (the die is bigger than 1" square), I'm sure an enterprising overclocker could thin a Pentium.
Moore's Law is one reason why software still stinks. Instead of perfecting systems within the confines of a limited amount of resources, its too easy to just assume more MHz, MB, amd Mbps.
With exponentially increasing resources, nothing ever stabilizes and everyone knows it. If people design software with the assumption that it will be totally obsolete and replaced in 18 months, they create software that is so badly designed that it must be replaced in 18 months.
Until hardware performance plateaus and people get off the upgrade-go-round, programming will be sloppy and ugly.
Uh oh... now you've got me worried. What happens when a script kiddie hacks into my fridge and orders a million gallons of ice cream in my name? I suppose if it's a Microsoft Fridge (tm), it's going to need frequent patching. Or I could use an Apple Macintosh Fridge, which will be more secure but hold only a few kinds of food.
LOL! And perhaps Linux fridge would only accept plain-text standardized food found for free on the road side?
But the script kiddie/Home appliance problem might be worse than you think. What happens when a script kiddie turns the oven on to 500 degrees (remote controlled for warming food) while you are out. Wiping out the home directory is one thing, burning down the home is another.
A few people earlier today were wondering why anyone would need IPv6, since IPv4 "obviously has enough address space". Developments like this should pretty clearly demonstrate that that's not the case. It probably won't be too terribly long before even your fridge will need an IP so you can program your refrigerator to know when it needs to order more groceries and the like.
Absolutely. In fact, your fridge might demand an entire subnet. Smaller, cheaper boards drive appliance makers to a federated, modular architectures in which every new function has its own CPU. Your fridge might need range of IPs addys if it has an ice maker, RFID-reading intelli-chiller, home-message center, Kalory-Kounter terahertz sensor array, Phreshness Gas Sensor, Open-Door SMS alert sender, remote shopping list VPN website, etc.
Its just much easier to make a bunch of modules that sit on a network than create a bloatware central system that has wires for every conceivable add-on function.
The downside of opening up the TLD system is the potential for gated communities that fragment the internet. Some TLDs might decide to only accept conections from particular other TLDs. They might do this to weed out spam, viruses, or objectionable material from other countries.
Some countries, like the US, could legislate that all pron and violent materials be relegated to particular TLDs that let parents easily filter out this material. Other countries might have similar rules or use content-category TLDs for censorship purposes.
Currenty, TLDs tell you nothing about the reputability of the domain owner -- anyone can get a domain at any TLD. Competition between TLDs could be a good thing in this regard. Some TLDs might become very selective of members -- creating TLDs with high reputations. This is in contrast to some domains, like.biz, that appear to be the lairs for so many spammer ecommerce sites (as far as I have seen).
It would be nice to be able to trust organizations that have a particular TLD -- knowing that the could not get and retain that TLD unless they adhered to a strict ethical code and had the organizational resources to support whatever products/services/info they were providing.
If everyone who owned a website posted thousands of bogus email addresses, then spammers harvesting efforts would quickly become useless. It should not be too hard to litter the web with billions of false e-mail addresses on bot-finadable pages.
If you really want to hurt the spam industry, then synthesize harvestable addresses based on the domains of e-commerce sites that use spam. If the mail server at cheapviagra.biz starts getting thousands or millions of emails from other spammers, its going to impact the ecommerce sites that use spam.
Those emails still go somewhere. And when my mail servers go nuts because there full of mails that won't relay due to the fact they won't resolve, i'll be looking for legs to break.
Unless you are relaying mail for spammers (not good) or are hosting pages with dummy addys (possible if you are ahosting service and your customers create dummy addys), your mail servers don't have to handle anything.
Spammers profit no matter how much mail they have to send, and no matter how many of those email addresses are bad. The bandwidth costs to send out hundreds of millions of emails is basically nil, compared to what they make back on sales to those poor people dumb enough to actually buy the products they're advertising.
Not true. While, spammers do make money at very low rates of return, reducing the rate of return would hurt them. If spammers get 1/10 or 1/100 the number of clickthroughs, they will feel that.
Even if spammers use zombies to send mail, that resource is finite. If spammers find they need one hundred times as many zombies to get the same number if idiots to buy their junk, it will impact them.
If everyone who owned a website posted thousands of bogus email addresses, then spammers harvesting efforts would quickly become useless. It should not be too hard to litter the web with billions of false e-mail addresses on bot-finadable pages.
The more enterprising site and mail server owners could even create semi-real bot email addresses that simply forward all emails to authorities. Even better, the mail server might first appear to "look at" spam by using an automated process to appear to fetching the coded JPGs that tell the spammer they have a live address. After the spammer thinks they have a good address, all further email would be sent directly to authorities.
This could be a DDoP (Distributed Denial of Profits) attack on harvesters and spammer. By creating ten to a hundred times the number of bad addresses as good addresses, we could reduce profit per spam by a factor of ten to a hundred and create a massive stream of data samples for authorities to use to catch spammers.
If car companies reveal enough of the interface design, then hackers will be able to disable or erase the car's black boxes. That way the police won't know if you were speeding when you wrap the car around a telephone pole.
My wife and I have worked together for 9 years in a small consulting business (we've been married 20 years). It works very well for us because we have complimentary skills, mutual respect, and agree on many issues of business straetgy and tactics. She can do things I can't do, and vice versa.
If you try to have a boss-employee relationship with your girlfiend or family, things might get ugly when you have to make an executive decision that they do not agree with or respect. You could try establishing "ground rules" but I'd bet that any asymmetries in the relationship, even if prearranged, will lead to grief.
This is a high-risk, high-reward issue. If you make this family business work, you will have the best time of your life. If you can't get along with family/coworkers you will have the worst time of your life.
This distribution method seems ideal for Linux also. Perhaps if HP weren't afraid of MS, we could also get nice bootable Linux distro while waiting for a venti mocha.
Why do people get so worked up over silly categories? Because "Man was the animal with the pigeon-hole brain," said my college biology professor.
The idea that there must be some hard cut-off between planets and non-planets seems ludicrious to me (IANAA). Its not unlike the taxonomy debates over species and genera -- artificial categories that don't stand up to the bizzare spectrum of possibilities that are out there. I'm sure when the solar system formed, it formed a wide range of objects of different sizes. The cut-off is just an arbitrary number that lets professors write papers and get tenure.
I will admit that categories are a very useful approximations for the world. Categories help us create knowledge at the Class level that can be inherited by all reputed instances of that category. But its only an approximation as the membership functions are often not binary.
Core weakness of PageRank
on
In Google We Trust
·
· Score: 4, Insightful
The basic concept of PageRank is flawed because it assumes a monotonic ordering of sites on some single scale (e.g., popularity as defiend by linkage). The problem with PageRank is not the use of links to assess popularity, but the presumption of a single scale.
The search of "Apple" illustrates this well. This search, like s many is deeply ambiguous. It could refer to the computer company, to the fruit, to the record company, to New York City, or to Apple Valley (MN or CA). Even if you know it refers to the company, its still ambiguous. It could refer to the company (as an investment), the products (for purchase), or a question(as in technical support).
The point is that each of these ambiguous alternatives creates an independent cluster of hits. Although one can create a ranking within each cluster, it is impossible to construct a meanful rank for all hits across all clusters - the second hit for "Apple computer" is not comparable to the 2nd hit for "Apple Records".
Instead of a pagerank scheme that sorts the universe of hits the instant the user enters the search, search engines should be more interactive. The first page of hits would emphasize breadth -- displaying hits most representative of their respective alternative clusters. As the searcher selects hits, the subsequent pages might show popularity-ranked hits within the clusters that seem to interest the searcher.
Each hit and each page would serve a double-duty -- serving the searcher's need to get information from the internet, and answering the search engine's question about the needs of the searcher. Until the search engine understands each searcher and each search, it cannot hope to rank the hits.
If these two outer gas giants cool further, I wonder if the convecting fluid layer will become thinner and thinner as the core freezes? If so, the magnetic field should split into more and more domains. Ultimately, the remaining thin layer will form a dense pattern of Rayleigh-Benard convection cells.
Those small domains will be hard to detect, though. As planet moves to equilibrium and the fluid layer thins and cools, the delta-T driving the convention will weaken. Smaller cells of slower-moving fluid will mean a much weaker magnetic field. Now we only need to wait a few billion years to see if this is what will happen......
I talked to a Bell executive a few years ago and he had an interesting stories about how the Bells created new technologies.
At the same time that fiberoptics were invented, Ma Bell had another high-bandwidth long-distance telecom technology in the works. Microwaves travelling in underground copper pipes could carry a modestly high bandwidth signal for long distances. They actually had an entire factory to creating the equipment (pipe, connectors, repeaters, edge boxes, etc.) When fiber came out (with its superior cost structure and tech performance) they simply killed the concept and the factory and adopted fiber.
Another Version of This Concept
on
The Power of Sewage
·
· Score: 4, Interesting
We discussed a similar high temperature conversion in the past. This alternative process uses high temperature/high pressure water to crack a wide range of complex molecules into simpler stuff. It can convert sewage, toxic waste, and animal byproducts into a mix of combustible hydrocarbons, salts, and water.
The new Microbial Fuel Cell method sounds interesting, but I bet it fails in the field. I'd bet that nasty substances (the odd pulse of heavy metals, detergents, or drain cleaner) would poison the microbial catalysts in this new fuel cell.
The problem, I think, is really an issue of lack of vision combined with a kind of greediness. We really can obtain utility computing... for some kinds of task. For others it's more difficult, maybe impossible. Modularity is (must be, IMHO) the answer.
Exactly! Modularity is the key to utility computing. And simple interfaces are the key to modularity. Very good points.
The challenge is to define a very clear, very simple (= restrictive) standard and ensure that all elements (both hardware and software) conform to that standard. In many domains this is not easy. Anything that requires a UI quickly becomes subject to the upgrade merry-go-round . Even in the case of strict numerical computing you can run into unexpected behavior if the processor you are now using in a grid does not have the same patterns of round-off error as the processor you used for development of the algorithms. This doesn't even get into dependencies with underlying math libraries or problems with machines that lack parity or ECC RAM.
I also wonder to what extent key networking protocols and kernel elements are unchanging enough -- would the same grid executable work and return the same results 20 years from now or do we have to upgrade the dispatcher and core software for each new OS and hardware architecture? (For example, I think that SETI has steadily migrated to newer and newer platforms to the point that it no longer runs on older hardware).
Overall, I definitely agree with you. Utility computing is possible for some domains and a simple modular design is the key to a stable, reliable, scalable architecture. And, as you said, the other key is to keep greedy people from trying to constantly "improve" things.
What's the primary requirement of a utility? It has to work. If you turn the water faucet on, you expect to get water; if you plug a lamp into a wall socket, you expect electricity; when you pick up your phone, you expect to hear a dial tone.
Exactly! But we will never reach this for three reasons.
1. Moore's Law: Ever expanding functional performacne levels means that computing does not stablize and we will never reach the reliability levels that would earn the name "utility computing." The problem is that faster clock sppeds and more transistors/chip permits an ever expanding set of standards, features, applications, and designs. In contrast, the U.S. has been using 120 VAC, 60 Hz for a 100 years with no real changes, upgrades, version 2 specs, etc.
2. Intercoupling of Devices & Applications: Unlike other utilties, computing elements are highly intercoupled through very complex interfaces. The "API" for electricity, water is simple -- what voltage/pressure and how much current/flow. Water and eletrical devices are mutually independent -- I don't have to upgrade all my lightbulbs when I get a new refrigerator. With traditional utilites, everything is truely plug and play because nothing interacts with anything else (with the minor exception of capacity limits). But with computing, each new feature, standard, operating system, and application has the potential to break other elements of my computing architecture. Until computing can get off the upgrade merry-go-round, it will not be a utility (see #1 for why that will not happen).
3. Industry Structure: The computing industry is structured very differently from most utilties. The independence of computer chip makers, PC makers, OS makers, application makers, and peripheral makers drives incompatibilities and unpleasant interactions between devices. Traditional utilities were highly regulated, vertically integrated organizations (less true today because of deregulation). The Bell System could introduce new features (e.g., PBX systems or touch tone phones) because they owned the phone, the phone making factory, the wires, the switching, the trunk lines, everything. In computing, if you have a problem each vendor can blame a different vendor for the problem and nothing really gets fixed.
I'm not suggesting that we repeal Moore's law, disconnect computers from each other, or regulate the industry. I'm only suggesting that computing is fundamentally different from the simple reliable utilties that "utility computing" would like to emulate.
Re:Prior Art: I know, RTFA (Impracticality?)
on
Pop Up Ads in Space
·
· Score: 3, Insightful
This is a very different device being claimed here, one that can project light down to earth, not just painting something on a rocket.
So true. Mod me Embarassed!
But now that I look at this, I wonder about its practicality. The mirror constellation will either need some very large mirrors to project sunlight over "intercontinental" distances or only work for small areas at a time. (a flat mirror in GEO would only create about a 200 mile diameter cone of visiblity on Earth)
Also, he will have a bit of a trade-off on the orbit for the system. LEO will put his satellites in Earth's shadow soon after dark (his sats will compete with dusk and then go dark). LEO is also hit-or-miss on whether the sats are flying over the target audience at exactly dusk (perhaps a resonant orbit would work). GEO provides better light and is stationary above the target audience, but the constellation will need to be much bigger (span hundreds of miles) and the mirrors much bigger to create a visible sign.
I think Pizza Hut has some prior art from 1999 on this one unless that Russian was behind the deal.
Security Threat of USB Flash Drives
on
USB Swiss Army Knife
·
· Score: 5, Insightful
I'd bet that devices like these give security officers, IP hoarders, and trade secret people the willies. The idea that someone (employee, vendor, or confident hacker) could walk into any office, stick their keychain USB drive into a PC and transfer files to/from an internal network is not too pleasant to contemplate. Seems like a great way to introduce trojans or snarf sensitive files. I wonder if some companies disable USB Mass Storage on their PCs to prevent this type of unauthorized access?
Based on the article and PieSpy site, it seems that PieSpy only finds the existence of a connection between members -- a symmetric relationship in which "A connects to B" implies "B connects to A". Yet human relationships tend to be asymmetric: "A likes B" does not imply that "B likes A" and "A controls B" certainly does not imply "B controls A".
A more powerful version of PieSpy would examine the text (and context) of who is connecting to whom. For example, the introduction of new words by some members of the network and the echoing of those words by others would help identify the directional flow of information in the network and help assess the level of control of the thread by some members over others. Analyzing the emotional content of words in threads could probably even let the software make approximate judgements of who likes/hates whom. Analyzing when some members leave IRC as a function of the joining of other members might also help detect asymmetric relationships.
Slashdot Mirror
Perhaps the problem is that the internet is too massively large for the human mind, social systems, and trademark laws to handle. Everyone thinks they are coming up with a unique, non-overlapping, name from everyone else. But once the system becomes too larger, very few names are unique. In reality, all the "good" names are taken and even all the easy variants of the good names are taken. Its a case of too many people and too few names.
The case of confusing/typoed near-names (ggle.com) is also a human scalability problem. If one only interacted within a tribe or small group (say 100 individuals), a typo or near-name would still be unambiguous.
People (and their social/legal systems) weren't designed to connect directly to millions or billions of others.
These techniques seem like brute force schemes to deal with the thermal resistance of chip packages -- you have to cool the heatsink to -110C in order to keep the "intel inside" at less than +60C). Why not use backside thinning. to bring the hot circuits of the processor within microns of a high coolant flux chamber. Backside thinning could get the coolant to within 10 microns of the junctions. If the CCD people can thin a massive 2k x 2k CCDs (the die is bigger than 1" square), I'm sure an enterprising overclocker could thin a Pentium.
Moore's Law is one reason why software still stinks. Instead of perfecting systems within the confines of a limited amount of resources, its too easy to just assume more MHz, MB, amd Mbps.
With exponentially increasing resources, nothing ever stabilizes and everyone knows it. If people design software with the assumption that it will be totally obsolete and replaced in 18 months, they create software that is so badly designed that it must be replaced in 18 months.
Until hardware performance plateaus and people get off the upgrade-go-round, programming will be sloppy and ugly.
Uh oh... now you've got me worried. What happens when a script kiddie hacks into my fridge and orders a million gallons of ice cream in my name? I suppose if it's a Microsoft Fridge (tm), it's going to need frequent patching. Or I could use an Apple Macintosh Fridge, which will be more secure but hold only a few kinds of food.
LOL! And perhaps Linux fridge would only accept plain-text standardized food found for free on the road side?
But the script kiddie/Home appliance problem might be worse than you think. What happens when a script kiddie turns the oven on to 500 degrees (remote controlled for warming food) while you are out. Wiping out the home directory is one thing, burning down the home is another.
A few people earlier today were wondering why anyone would need IPv6, since IPv4 "obviously has enough address space". Developments like this should pretty clearly demonstrate that that's not the case. It probably won't be too terribly long before even your fridge will need an IP so you can program your refrigerator to know when it needs to order more groceries and the like.
Absolutely. In fact, your fridge might demand an entire subnet. Smaller, cheaper boards drive appliance makers to a federated, modular architectures in which every new function has its own CPU. Your fridge might need range of IPs addys if it has an ice maker, RFID-reading intelli-chiller, home-message center, Kalory-Kounter terahertz sensor array, Phreshness Gas Sensor, Open-Door SMS alert sender, remote shopping list VPN website, etc.
Its just much easier to make a bunch of modules that sit on a network than create a bloatware central system that has wires for every conceivable add-on function.
The downside of opening up the TLD system is the potential for gated communities that fragment the internet. Some TLDs might decide to only accept conections from particular other TLDs. They might do this to weed out spam, viruses, or objectionable material from other countries.
Some countries, like the US, could legislate that all pron and violent materials be relegated to particular TLDs that let parents easily filter out this material. Other countries might have similar rules or use content-category TLDs for censorship purposes.
Currenty, TLDs tell you nothing about the reputability of the domain owner -- anyone can get a domain at any TLD. Competition between TLDs could be a good thing in this regard. Some TLDs might become very selective of members -- creating TLDs with high reputations. This is in contrast to some domains, like .biz, that appear to be the lairs for so many spammer ecommerce sites (as far as I have seen).
It would be nice to be able to trust organizations that have a particular TLD -- knowing that the could not get and retain that TLD unless they adhered to a strict ethical code and had the organizational resources to support whatever products/services/info they were providing.
If everyone who owned a website posted thousands of bogus email addresses, then spammers harvesting efforts would quickly become useless. It should not be too hard to litter the web with billions of false e-mail addresses on bot-finadable pages.
If you really want to hurt the spam industry, then synthesize harvestable addresses based on the domains of e-commerce sites that use spam. If the mail server at cheapviagra.biz starts getting thousands or millions of emails from other spammers, its going to impact the ecommerce sites that use spam.
Lets use spammers to spam other spammers.
Those emails still go somewhere. And when my mail servers go nuts because there full of mails that won't relay due to the fact they won't resolve, i'll be looking for legs to break.
Unless you are relaying mail for spammers (not good) or are hosting pages with dummy addys (possible if you are ahosting service and your customers create dummy addys), your mail servers don't have to handle anything.
Spammers profit no matter how much mail they have to send, and no matter how many of those email addresses are bad. The bandwidth costs to send out hundreds of millions of emails is basically nil, compared to what they make back on sales to those poor people dumb enough to actually buy the products they're advertising.
Not true. While, spammers do make money at very low rates of return, reducing the rate of return would hurt them. If spammers get 1/10 or 1/100 the number of clickthroughs, they will feel that.
Even if spammers use zombies to send mail, that resource is finite. If spammers find they need one hundred times as many zombies to get the same number if idiots to buy their junk, it will impact them.
If everyone who owned a website posted thousands of bogus email addresses, then spammers harvesting efforts would quickly become useless. It should not be too hard to litter the web with billions of false e-mail addresses on bot-finadable pages.
The more enterprising site and mail server owners could even create semi-real bot email addresses that simply forward all emails to authorities. Even better, the mail server might first appear to "look at" spam by using an automated process to appear to fetching the coded JPGs that tell the spammer they have a live address. After the spammer thinks they have a good address, all further email would be sent directly to authorities.
This could be a DDoP (Distributed Denial of Profits) attack on harvesters and spammer. By creating ten to a hundred times the number of bad addresses as good addresses, we could reduce profit per spam by a factor of ten to a hundred and create a massive stream of data samples for authorities to use to catch spammers.
If car companies reveal enough of the interface design, then hackers will be able to disable or erase the car's black boxes. That way the police won't know if you were speeding when you wrap the car around a telephone pole.
My wife and I have worked together for 9 years in a small consulting business (we've been married 20 years). It works very well for us because we have complimentary skills, mutual respect, and agree on many issues of business straetgy and tactics. She can do things I can't do, and vice versa.
If you try to have a boss-employee relationship with your girlfiend or family, things might get ugly when you have to make an executive decision that they do not agree with or respect. You could try establishing "ground rules" but I'd bet that any asymmetries in the relationship, even if prearranged, will lead to grief.
This is a high-risk, high-reward issue. If you make this family business work, you will have the best time of your life. If you can't get along with family/coworkers you will have the worst time of your life.
Good Luck!
This distribution method seems ideal for Linux also. Perhaps if HP weren't afraid of MS, we could also get nice bootable Linux distro while waiting for a venti mocha.
Why do people get so worked up over silly categories? Because "Man was the animal with the pigeon-hole brain," said my college biology professor.
The idea that there must be some hard cut-off between planets and non-planets seems ludicrious to me (IANAA). Its not unlike the taxonomy debates over species and genera -- artificial categories that don't stand up to the bizzare spectrum of possibilities that are out there. I'm sure when the solar system formed, it formed a wide range of objects of different sizes. The cut-off is just an arbitrary number that lets professors write papers and get tenure.
I will admit that categories are a very useful approximations for the world. Categories help us create knowledge at the Class level that can be inherited by all reputed instances of that category. But its only an approximation as the membership functions are often not binary.
The basic concept of PageRank is flawed because it assumes a monotonic ordering of sites on some single scale (e.g., popularity as defiend by linkage). The problem with PageRank is not the use of links to assess popularity, but the presumption of a single scale.
The search of "Apple" illustrates this well. This search, like s many is deeply ambiguous. It could refer to the computer company, to the fruit, to the record company, to New York City, or to Apple Valley (MN or CA). Even if you know it refers to the company, its still ambiguous. It could refer to the company (as an investment), the products (for purchase), or a question(as in technical support).
The point is that each of these ambiguous alternatives creates an independent cluster of hits. Although one can create a ranking within each cluster, it is impossible to construct a meanful rank for all hits across all clusters - the second hit for "Apple computer" is not comparable to the 2nd hit for "Apple Records".
Instead of a pagerank scheme that sorts the universe of hits the instant the user enters the search, search engines should be more interactive. The first page of hits would emphasize breadth -- displaying hits most representative of their respective alternative clusters. As the searcher selects hits, the subsequent pages might show popularity-ranked hits within the clusters that seem to interest the searcher.
Each hit and each page would serve a double-duty -- serving the searcher's need to get information from the internet, and answering the search engine's question about the needs of the searcher. Until the search engine understands each searcher and each search, it cannot hope to rank the hits.
If these two outer gas giants cool further, I wonder if the convecting fluid layer will become thinner and thinner as the core freezes? If so, the magnetic field should split into more and more domains. Ultimately, the remaining thin layer will form a dense pattern of Rayleigh-Benard convection cells.
Those small domains will be hard to detect, though. As planet moves to equilibrium and the fluid layer thins and cools, the delta-T driving the convention will weaken. Smaller cells of slower-moving fluid will mean a much weaker magnetic field. Now we only need to wait a few billion years to see if this is what will happen......
I talked to a Bell executive a few years ago and he had an interesting stories about how the Bells created new technologies.
At the same time that fiberoptics were invented, Ma Bell had another high-bandwidth long-distance telecom technology in the works. Microwaves travelling in underground copper pipes could carry a modestly high bandwidth signal for long distances. They actually had an entire factory to creating the equipment (pipe, connectors, repeaters, edge boxes, etc.) When fiber came out (with its superior cost structure and tech performance) they simply killed the concept and the factory and adopted fiber.
We discussed a similar high temperature conversion in the past. This alternative process uses high temperature/high pressure water to crack a wide range of complex molecules into simpler stuff. It can convert sewage, toxic waste, and animal byproducts into a mix of combustible hydrocarbons, salts, and water.
The new Microbial Fuel Cell method sounds interesting, but I bet it fails in the field. I'd bet that nasty substances (the odd pulse of heavy metals, detergents, or drain cleaner) would poison the microbial catalysts in this new fuel cell.
The problem, I think, is really an issue of lack of vision combined with a kind of greediness. We really can obtain utility computing... for some kinds of task. For others it's more difficult, maybe impossible. Modularity is (must be, IMHO) the answer.
Exactly! Modularity is the key to utility computing. And simple interfaces are the key to modularity. Very good points.
The challenge is to define a very clear, very simple (= restrictive) standard and ensure that all elements (both hardware and software) conform to that standard. In many domains this is not easy. Anything that requires a UI quickly becomes subject to the upgrade merry-go-round . Even in the case of strict numerical computing you can run into unexpected behavior if the processor you are now using in a grid does not have the same patterns of round-off error as the processor you used for development of the algorithms. This doesn't even get into dependencies with underlying math libraries or problems with machines that lack parity or ECC RAM.
I also wonder to what extent key networking protocols and kernel elements are unchanging enough -- would the same grid executable work and return the same results 20 years from now or do we have to upgrade the dispatcher and core software for each new OS and hardware architecture? (For example, I think that SETI has steadily migrated to newer and newer platforms to the point that it no longer runs on older hardware).
Overall, I definitely agree with you. Utility computing is possible for some domains and a simple modular design is the key to a stable, reliable, scalable architecture. And, as you said, the other key is to keep greedy people from trying to constantly "improve" things.
What's the primary requirement of a utility? It has to work. If you turn the water faucet on, you expect to get water; if you plug a lamp into a wall socket, you expect electricity; when you pick up your phone, you expect to hear a dial tone.
Exactly! But we will never reach this for three reasons.
1. Moore's Law: Ever expanding functional performacne levels means that computing does not stablize and we will never reach the reliability levels that would earn the name "utility computing." The problem is that faster clock sppeds and more transistors/chip permits an ever expanding set of standards, features, applications, and designs. In contrast, the U.S. has been using 120 VAC, 60 Hz for a 100 years with no real changes, upgrades, version 2 specs, etc.
2. Intercoupling of Devices & Applications: Unlike other utilties, computing elements are highly intercoupled through very complex interfaces. The "API" for electricity, water is simple -- what voltage/pressure and how much current/flow. Water and eletrical devices are mutually independent -- I don't have to upgrade all my lightbulbs when I get a new refrigerator. With traditional utilites, everything is truely plug and play because nothing interacts with anything else (with the minor exception of capacity limits). But with computing, each new feature, standard, operating system, and application has the potential to break other elements of my computing architecture. Until computing can get off the upgrade merry-go-round, it will not be a utility (see #1 for why that will not happen).
3. Industry Structure: The computing industry is structured very differently from most utilties. The independence of computer chip makers, PC makers, OS makers, application makers, and peripheral makers drives incompatibilities and unpleasant interactions between devices. Traditional utilities were highly regulated, vertically integrated organizations (less true today because of deregulation). The Bell System could introduce new features (e.g., PBX systems or touch tone phones) because they owned the phone, the phone making factory, the wires, the switching, the trunk lines, everything. In computing, if you have a problem each vendor can blame a different vendor for the problem and nothing really gets fixed.
I'm not suggesting that we repeal Moore's law, disconnect computers from each other, or regulate the industry. I'm only suggesting that computing is fundamentally different from the simple reliable utilties that "utility computing" would like to emulate.
This is a very different device being claimed here, one that can project light down to earth, not just painting something on a rocket.
So true. Mod me Embarassed!
But now that I look at this, I wonder about its practicality. The mirror constellation will either need some very large mirrors to project sunlight over "intercontinental" distances or only work for small areas at a time. (a flat mirror in GEO would only create about a 200 mile diameter cone of visiblity on Earth)
Also, he will have a bit of a trade-off on the orbit for the system. LEO will put his satellites in Earth's shadow soon after dark (his sats will compete with dusk and then go dark). LEO is also hit-or-miss on whether the sats are flying over the target audience at exactly dusk (perhaps a resonant orbit would work). GEO provides better light and is stationary above the target audience, but the constellation will need to be much bigger (span hundreds of miles) and the mirrors much bigger to create a visible sign.
I think Pizza Hut has some prior art from 1999 on this one unless that Russian was behind the deal.
I'd bet that devices like these give security officers, IP hoarders, and trade secret people the willies. The idea that someone (employee, vendor, or confident hacker) could walk into any office, stick their keychain USB drive into a PC and transfer files to/from an internal network is not too pleasant to contemplate. Seems like a great way to introduce trojans or snarf sensitive files. I wonder if some companies disable USB Mass Storage on their PCs to prevent this type of unauthorized access?
Based on the article and PieSpy site, it seems that PieSpy only finds the existence of a connection between members -- a symmetric relationship in which "A connects to B" implies "B connects to A". Yet human relationships tend to be asymmetric: "A likes B" does not imply that "B likes A" and "A controls B" certainly does not imply "B controls A".
A more powerful version of PieSpy would examine the text (and context) of who is connecting to whom. For example, the introduction of new words by some members of the network and the echoing of those words by others would help identify the directional flow of information in the network and help assess the level of control of the thread by some members over others. Analyzing the emotional content of words in threads could probably even let the software make approximate judgements of who likes/hates whom. Analyzing when some members leave IRC as a function of the joining of other members might also help detect asymmetric relationships.