One of the main questions here is whether we could do a better job with the Windows, x86, JCL, etc. if we started over. It's possible that we could move everything to quantum RISC chips, but at the same time we would probably incur the same or worse "industry standards" as we have now.
For one thing, we can't assume that all the people who use BASIC will perish. At best we can hope that they'll end up on an ice floe somewhere, chipping GOTO's into whale bones. But the fact is, at some point, someone will say "hey, we can get BASIC to run on this", and the game will be up. (I don't mean to disparage BASIC too much; I wrote stuff in it too, but for the sake of argument let's assume that it's undesirable).
But more importantly, the process of putting together standards would be even more fragmented. Most of the gobbledygook in technology arises from copying functionality from older platforms. This is a process of social replication, that is, people copy things that they know and obtain from other people they know. This same process has been observed in other knowledge-based systems like the web and in patents and academic journals - people don't make *random* links between ideas, web pages, etc., they tend to copy them from other sources, and make small modifications.
This process depends on communication. By discussing or searching for other people's solutions to our problems, we can avoid duplication of effort. However, if we presumably have lost our computers and communications networks, we won't have the chance to unite to solve particular problems. Many competing standards will arise in areas that are isolated from each other. The ability for a competitive marketplace to weed out the crap will be heavily impaired. At best we will be left with small "communities" which are far more isolated than the current Linux or Mainframe or RISC communities. The process of uniting these groups will inevitably have the same problems we have now - incompatible solutions to similar problems, lack of interest in, or hostility to, common standards, and fights over patents and intellectual property, whether real or imagined.
Also, you could pass the signal through one of those light-speed-reduced materials that people keep coming up with. I'm not sure what it would do to the signal quality, but you could get a heck of a lot of data into a pipe where light goes 42 mph.
The method relies on the random number stream not being bufferable for the time it takes to crack the start-time message.
However the speed of light makes it easy to delay the RNG signal by reflecting it off of a suitably distant relay station a few light-minutes or hours away, say near Mars. By the time the signal makes it back, the start time can be cracked and ready for decryption.
The situation in Korea is that there are thousands of "PC-pang"'s which are basically pay-by-the-hour PC facilities. I would imagine that this would be a perfect situation for a NAT, although I would expect the ISP's to take that into account in a commercial DSL subscription.
Companies that wire up buildings there work on a revenue-share basis with ISP's, so there's no incentive to let tenants share a line or work around the provider.
Possibly the PC-pangs are getting dirt cheap residential service and NATing it to 10-20 active PC's. I can see some reason for complaint with that, but that's more of a commercial vs. residential service fraud issue.
Use google's cache. http://www.google.com/search?q=cache:www.domain.co m will bring up the cached version of the page (with the google header). They can probably tweak this to avoid abuse, but for now it's open.
I think a lot of time is wasted waiting for pages to appear. I routinely use google's cached version of a page to avoid indefinite waits (and in fact the "stale" version is often better than a page that's gone dead - after all it's more likely to match my query). Large-scale caching a la Akamai is probably the way to go for a lot of sites.
It would be interesting to build a distributed network of cache sites to decrease latency. My thought is to use a proxy of some sort to get requested pages, and then stream them out on IP multicast to other proxies. Each client proxy could then maintain a certain number of recently-broadcast pages as a cache. I think this has a lot of similarity to Gnutella, so maybe a gateway could be built around this program instead.
For the record, the Golden Gate Bridge District has never put safety, or the letter of the law, at the top of its priorities, and it has been negligent for the past decade.
In 1990, a man was riding a bicycle across the bridge when a pedestrian stepped in front of him. The resulting collision caused the bicyclist to be thrown over the ankle-high "safety railing" that separates car traffic from pedestrians and bicyclists, and he was quickly run over and killed. The bridge was deemed negligent if it failed to correct the problem, but in ten years, nothing has been done. They put six foot temporary fences along the railing to protect their workers from falling tools, but a fence half as high along the inside railing has never been attempted.
A few years ago, a two-year old slid through the gap beneath this same railing and died from the resulting fall. The bridge district immediately placed a cable across the openings. However the height of the barriers is still the same. Perhaps if they had fixed the barriers as they should have originally, the child might not have died.
Interesting note: This boat seems to have the same series engine as the Boeing 777, the Rolls-Royce Trent. I don't know how these compare in actual fuel economy, but the ship carries 10,000 tons with five engines, while (I think) the plane carries only 20 tons with four.
But that doesn't make sense - the 777 carries up to 550 passengers. At.1 ton each, that makes 55 tons. I think the figures I have are for passenger planes. Maybe the total cargo capacity is close to 100 tons; that's still only one percent of the ship's capacity. If you factor in a 10x speed advantage for the plane, you still only get 10 percent of the ship's capacity for roughly the same fuel usage.
According to some dinner conversation I had a few years ago, Korean Air ships DRAM into SFO by plane, and ships back cherries on the same planes. Kind of a weird juxtaposition.
I'm from California, and I can't really disagree with you, but I think the flip side of this deregulation is that PG&E is now bidding on *your* power, all over North America. That's how the wholesale market is supposed to operate. So when Bush says he "sees a ripple effect" that's idiot-speak for "your local power company is selling kilowatt-hours on ebay". After all, does your utility want to sell you power for 10 cents a kwh, or sell to PG&E for 30?
Of course, the solution is to drill for more oil, not to conserve, according to President Clampett.
Here's a cool invention for recovering heat from hot water going down the drain. Apparently the way water runs down the inside of a large-diameter sewer pipe is actually pretty good for heat exchange, so you don't need a complicated manifold to get good efficiency. Another good feature is that it reduces the size of the tank needed in the water heater, since peak loads like showers will also produce lots of recycled heat.
You might also try just taking baths and letting the water cool in the tub before you drain it. I haven't calculated, but I suspect you could get back a lot of energy this way.
Washers and dryers also tend to waste a lot of heat by dumping hot stuff without recovering heat. There are some new designs for "heat pump dryers" (hint: put into search engine) which condense the evaporated water and pump the heat back into the clothes. You could also build a heat exchanger that would cool the dryer output and bring the heat indoors; you just need to handle the lint and runoff.
This technology was hot for a while, and eventually fizzled out for the usual one-size-doesn't-fit-all reasons, but there are a couple of features in it that I like:
1. Cross-referenced data elements and pseudocode. This means that any time you build a statement like "put tab A in slot B" then tab A and slot B are already defined in the data model. In fact just about all the nouns and relationships (eg "find all where is ") should be pickable from the data model (the items in above should all be such). Even the types of local variables should be pickable from the model.
2. In keeping with the above, you need a rich data model. That means you have to give relationships names and correct cardinalities, and put in things like delete rules for foreign keys. Too many times I see "data modellers" slap together a few entities with unnamed lines for the relationships, and then end up with a denormalized mess because no one could figure out what the relationships mean. Incorrect or misunderstood relationships are where most bugs come from. Keeping the data model up-to-date can be a real pain, but the alternative is chaos.
One of the biggest power drains, in the Bay Area at least, is the wind tunnels at NASA Ames in Mountain View. These have their own substation, and run at night to avoid the peak. Imagine an 80x120 foot fan, blowing at 200 mph, or pumping a big tunnel up to several atmospheres - that's a lot of energy (particularly if it explodes, which is why the high-pressure tunnel was shut down).
Maybe now that power is going up to 30 cents per kwh they can test the planes with their engines on, and use the fans as generators. I doubt if they'll do much aeronautical research.
There are two political issues colliding here, which cause some confusion. First is the recent power crisis, which is the result of flawed deregulation, which is (seen as) the result of PG&E lobbying. Second is the longstanding problem in accounting for the output of the Hetch-Hetchy reservoir. It seems that San Francisco acquired the rights to the output of this reservoir 80-90 years ago, but, for some reason, city residents have always had to pay PG&E for their power, while, somehow, PG&E has been reselling the output from the dam to other customers.
Given the fact that PG&E has its headquarters in SF, and the general sorry state of the city's politics, it's easy to conclude that something shady went on here.
For San Francisco, it seems the easiest way out of the power crisis is to reassert ownership of the Hetch Hetchy output and get PG&E out of it. The dam was built to supply power to the city at a fixed cost indefinitely, and being forced to pay spot rates for the same juice is ridiculous.
Sorry, I was making a broad stereotype to get the point across.
No, it's true. Sometime in the 70's, there was a mass exodus of chainsaw-toting, fur-coat-wearing nuclear engineers who were forced to drop their Big Macs and run for the Nevada border, towards Area 51.
You can time things by observing Jupiter's moons, or any other well-known astronomic event. Clocks were developed for seaborne use due to the difficulty of using a telescope on board, but the astronomical method is still useful on land. Basically if you know that everyone in the world is looking at the same event at the same time, you can compare measurements at that instant, say between Greenwich and your location, and get the longitude difference from that.
In San Francisco we just bought a bunch of Bredas from Italy (what part I don't know). They're great.
What the author didn't mention is that SF is expanding its light rail system with as much money as it can muster. In the last decade we've added new (or restored) track on Market Street and the Embarcadero, and we're planning a major new line down Third Street and possibly up the Geary corridor. It's one of the fastest-growing parts of the SF transportation system.
There are still plenty of idiots who think we should bulldoze $500M worth of housing and put in freeways,etc., but most of them live down in San Jose or LA, where people do those kinds of things;-)
APL has some unique ideas that I haven't seen in other languages. One of the most powerful is to allow arrays of arbitrary dimension to be combined or processed using the same operators that apply to scalars. For instance, "A+B" is the same whether A and B are integers, matrices, or 3x10x4 arrays. In either case the result is another object of the same dimension, which can be fed into another operator, and so on.
Other operators allow collapsing or aggregating along one dimension, eg +/A sums up A along its "first" dimension and produces an array of the results. Another operator transposes the array to put the dimensions in the order needed to match other arrays, etc. By transposing and aggregating, complicated relational operations can be performed in one line of code.
In fact, I realized recently that APL has many of the capabilities of SQL, but it is many times more concise. The reason is that, while SQL demands writing out join conditions for every "dimension" (join key) in an expression, APL works by knowing which dimensions are supposed to match in each operation. For example, say A and B are "tables" with 3 "keys" (dimensions) d1-d3, similar to A and B above. If we want to add up A+B elementwise, we have to write out something like:
select A.d1, A.d2, A.d3, A.value + B.value
from A, B
where A.d1 = B.d1
and A.d2 = B.d2
and A.d3 = B.d3
This is cumbersome and varies according to the number of key fields in A and B. Hopefully you can see that APL is a better solution in this case.
Someday I hope to come up with an APL-like shorthand for doing relational queries. The main idea would be to have joins, cartesian products, aggregations, operators, etc. that work by implicitly matching keys rather than by listing them. In fact, a lot of the APL operators would work if applied directly to tables, eg
I can just see it now - a black van pulls up at the Optimum Online office, the side door slides open, and several black-clad men jump out, and do what? Take cover, and start returning fire?
You have to understand the frustration of running a search engine: you never know how good your result ranking is, because you never know which result the user clicked on, or whether the user simply gave up and went to another search engine.
The only way to track this behavior is to put up a redirect (or run an ISP - the free ones log just about everything).
Sorry I can't offer much insight, since I haven't read the book, but IIRC another review pegged this as David Bohm's story in disguise.
Bohm was a Princeton prof who revived de Broglie's Pilot Wave formulation, showed that it made sense, and then at some point was McCarthyized and ended up in Britain, blacklisted from US universities.
In any case, I like Bohm. He stacks up as follows:
Reality is:
deterministic, nonlocal, nonjumpy: Bohm, de Broglie
nondeterministic, local but jumpy: Bohr, Heisenberg, most basic physics books
deterministic, local: Einstein, Simpson (Homer)
nondeterministic, nonlocal, solipsistic: New Age physics-interpretation gurus
(sorry, it's Friday, hard to dredge this stuff up off the cuff)
Another random note: Most people don't know that de Broglie actually came up with his wave theory of matter to reconcile special relativity and the primitive quantum ideas of the time. The waves were a compensating factor for the slowdown of "internal vibrations" due to SR as v approaches c. So, in effect, QM is already a reconciliation with SR. However he eventually discarded his "pilot wave" theory, not because it didn't work, but because it didn't fit his assumptions about reality.
For more info on Bohmian Mechanics, here is a link to some current research.
Slashdot Mirror
One of the main questions here is whether we could do a better job with the Windows, x86, JCL, etc. if we started over. It's possible that we could move everything to quantum RISC chips, but at the same time we would probably incur the same or worse "industry standards" as we have now.
For one thing, we can't assume that all the people who use BASIC will perish. At best we can hope that they'll end up on an ice floe somewhere, chipping GOTO's into whale bones. But the fact is, at some point, someone will say "hey, we can get BASIC to run on this", and the game will be up. (I don't mean to disparage BASIC too much; I wrote stuff in it too, but for the sake of argument let's assume that it's undesirable).
But more importantly, the process of putting together standards would be even more fragmented. Most of the gobbledygook in technology arises from copying functionality from older platforms. This is a process of social replication, that is, people copy things that they know and obtain from other people they know. This same process has been observed in other knowledge-based systems like the web and in patents and academic journals - people don't make *random* links between ideas, web pages, etc., they tend to copy them from other sources, and make small modifications.
This process depends on communication. By discussing or searching for other people's solutions to our problems, we can avoid duplication of effort. However, if we presumably have lost our computers and communications networks, we won't have the chance to unite to solve particular problems. Many competing standards will arise in areas that are isolated from each other. The ability for a competitive marketplace to weed out the crap will be heavily impaired. At best we will be left with small "communities" which are far more isolated than the current Linux or Mainframe or RISC communities. The process of uniting these groups will inevitably have the same problems we have now - incompatible solutions to similar problems, lack of interest in, or hostility to, common standards, and fights over patents and intellectual property, whether real or imagined.
(because I just posted the same idea).
But you beat me by 22 minutes.
Also, you could pass the signal through one of those light-speed-reduced materials that people keep coming up with. I'm not sure what it would do to the signal quality, but you could get a heck of a lot of data into a pipe where light goes 42 mph.
The method relies on the random number stream not being bufferable for the time it takes to crack the start-time message.
However the speed of light makes it easy to delay the RNG signal by reflecting it off of a suitably distant relay station a few light-minutes or hours away, say near Mars. By the time the signal makes it back, the start time can be cracked and ready for decryption.
The situation in Korea is that there are thousands of "PC-pang"'s which are basically pay-by-the-hour PC facilities. I would imagine that this would be a perfect situation for a NAT, although I would expect the ISP's to take that into account in a commercial DSL subscription.
Companies that wire up buildings there work on a revenue-share basis with ISP's, so there's no incentive to let tenants share a line or work around the provider.
Possibly the PC-pangs are getting dirt cheap residential service and NATing it to 10-20 active PC's. I can see some reason for complaint with that, but that's more of a commercial vs. residential service fraud issue.
Use google's cache. http://www.google.com/search?q=cache:www.domain.co m will bring up the cached version of the page (with the google header). They can probably tweak this to avoid abuse, but for now it's open.
I think a lot of time is wasted waiting for pages to appear. I routinely use google's cached version of a page to avoid indefinite waits (and in fact the "stale" version is often better than a page that's gone dead - after all it's more likely to match my query). Large-scale caching a la Akamai is probably the way to go for a lot of sites.
It would be interesting to build a distributed network of cache sites to decrease latency. My thought is to use a proxy of some sort to get requested pages, and then stream them out on IP multicast to other proxies. Each client proxy could then maintain a certain number of recently-broadcast pages as a cache. I think this has a lot of similarity to Gnutella, so maybe a gateway could be built around this program instead.
For the record, the Golden Gate Bridge District has never put safety, or the letter of the law, at the top of its priorities, and it has been negligent for the past decade.
In 1990, a man was riding a bicycle across the bridge when a pedestrian stepped in front of him. The resulting collision caused the bicyclist to be thrown over the ankle-high "safety railing" that separates car traffic from pedestrians and bicyclists, and he was quickly run over and killed. The bridge was deemed negligent if it failed to correct the problem, but in ten years, nothing has been done. They put six foot temporary fences along the railing to protect their workers from falling tools, but a fence half as high along the inside railing has never been attempted.
A few years ago, a two-year old slid through the gap beneath this same railing and died from the resulting fall. The bridge district immediately placed a cable across the openings. However the height of the barriers is still the same. Perhaps if they had fixed the barriers as they should have originally, the child might not have died.
Interesting note: This boat seems to have the same series engine as the Boeing 777, the Rolls-Royce Trent. I don't know how these compare in actual fuel economy, but the ship carries 10,000 tons with five engines, while (I think) the plane carries only 20 tons with four.
.1 ton each, that makes 55 tons. I think the figures I have are for passenger planes. Maybe the total cargo capacity is close to 100 tons; that's still only one percent of the ship's capacity. If you factor in a 10x speed advantage for the plane, you still only get 10 percent of the ship's capacity for roughly the same fuel usage.
But that doesn't make sense - the 777 carries up to 550 passengers. At
According to some dinner conversation I had a few years ago, Korean Air ships DRAM into SFO by plane, and ships back cherries on the same planes. Kind of a weird juxtaposition.
I'm from California, and I can't really disagree with you, but I think the flip side of this deregulation is that PG&E is now bidding on *your* power, all over North America. That's how the wholesale market is supposed to operate. So when Bush says he "sees a ripple effect" that's idiot-speak for "your local power company is selling kilowatt-hours on ebay". After all, does your utility want to sell you power for 10 cents a kwh, or sell to PG&E for 30?
Of course, the solution is to drill for more oil, not to conserve, according to President Clampett.
Maybe he's draining the oil out and burning it.
Oops, I used angle brackets and parts of my example got stripped out. I'll try brackets (and previewing for a change):
The bracketed items would be from the data dictionary...at least you can see how little is left without them(!).
Most lights are near the ceiling, and the hot air generated stays up there. Try to encourage your downstairs neighbors to use incandescents.
Here's a cool invention for recovering heat from hot water going down the drain. Apparently the way water runs down the inside of a large-diameter sewer pipe is actually pretty good for heat exchange, so you don't need a complicated manifold to get good efficiency. Another good feature is that it reduces the size of the tank needed in the water heater, since peak loads like showers will also produce lots of recycled heat.
You might also try just taking baths and letting the water cool in the tub before you drain it. I haven't calculated, but I suspect you could get back a lot of energy this way.
Washers and dryers also tend to waste a lot of heat by dumping hot stuff without recovering heat. There are some new designs for "heat pump dryers" (hint: put into search engine) which condense the evaporated water and pump the heat back into the clothes. You could also build a heat exchanger that would cool the dryer output and bring the heat indoors; you just need to handle the lint and runoff.
This technology was hot for a while, and eventually fizzled out for the usual one-size-doesn't-fit-all reasons, but there are a couple of features in it that I like:
1. Cross-referenced data elements and pseudocode. This means that any time you build a statement like "put tab A in slot B" then tab A and slot B are already defined in the data model. In fact just about all the nouns and relationships (eg "find all where is ") should be pickable from the data model (the items in above should all be such). Even the types of local variables should be pickable from the model.
2. In keeping with the above, you need a rich data model. That means you have to give relationships names and correct cardinalities, and put in things like delete rules for foreign keys. Too many times I see "data modellers" slap together a few entities with unnamed lines for the relationships, and then end up with a denormalized mess because no one could figure out what the relationships mean. Incorrect or misunderstood relationships are where most bugs come from. Keeping the data model up-to-date can be a real pain, but the alternative is chaos.
Good, then maybe you can supply the power to pump our water down there :-)
One of the biggest power drains, in the Bay Area at least, is the wind tunnels at NASA Ames in Mountain View. These have their own substation, and run at night to avoid the peak. Imagine an 80x120 foot fan, blowing at 200 mph, or pumping a big tunnel up to several atmospheres - that's a lot of energy (particularly if it explodes, which is why the high-pressure tunnel was shut down).
Maybe now that power is going up to 30 cents per kwh they can test the planes with their engines on, and use the fans as generators. I doubt if they'll do much aeronautical research.
There are two political issues colliding here, which cause some confusion. First is the recent power crisis, which is the result of flawed deregulation, which is (seen as) the result of PG&E lobbying. Second is the longstanding problem in accounting for the output of the Hetch-Hetchy reservoir. It seems that San Francisco acquired the rights to the output of this reservoir 80-90 years ago, but, for some reason, city residents have always had to pay PG&E for their power, while, somehow, PG&E has been reselling the output from the dam to other customers.
Given the fact that PG&E has its headquarters in SF, and the general sorry state of the city's politics, it's easy to conclude that something shady went on here.
For San Francisco, it seems the easiest way out of the power crisis is to reassert ownership of the Hetch Hetchy output and get PG&E out of it. The dam was built to supply power to the city at a fixed cost indefinitely, and being forced to pay spot rates for the same juice is ridiculous.
Sorry, I was making a broad stereotype to get the point across.
No, it's true. Sometime in the 70's, there was a mass exodus of chainsaw-toting, fur-coat-wearing nuclear engineers who were forced to drop their Big Macs and run for the Nevada border, towards Area 51.
You can time things by observing Jupiter's moons, or any other well-known astronomic event. Clocks were developed for seaborne use due to the difficulty of using a telescope on board, but the astronomical method is still useful on land. Basically if you know that everyone in the world is looking at the same event at the same time, you can compare measurements at that instant, say between Greenwich and your location, and get the longitude difference from that.
In San Francisco we just bought a bunch of Bredas from Italy (what part I don't know). They're great.
;-)
What the author didn't mention is that SF is expanding its light rail system with as much money as it can muster. In the last decade we've added new (or restored) track on Market Street and the Embarcadero, and we're planning a major new line down Third Street and possibly up the Geary corridor. It's one of the fastest-growing parts of the SF transportation system.
There are still plenty of idiots who think we should bulldoze $500M worth of housing and put in freeways,etc., but most of them live down in San Jose or LA, where people do those kinds of things
Other operators allow collapsing or aggregating along one dimension, eg +/A sums up A along its "first" dimension and produces an array of the results. Another operator transposes the array to put the dimensions in the order needed to match other arrays, etc. By transposing and aggregating, complicated relational operations can be performed in one line of code.
In fact, I realized recently that APL has many of the capabilities of SQL, but it is many times more concise. The reason is that, while SQL demands writing out join conditions for every "dimension" (join key) in an expression, APL works by knowing which dimensions are supposed to match in each operation. For example, say A and B are "tables" with 3 "keys" (dimensions) d1-d3, similar to A and B above. If we want to add up A+B elementwise, we have to write out something like:
This is cumbersome and varies according to the number of key fields in A and B. Hopefully you can see that APL is a better solution in this case.Someday I hope to come up with an APL-like shorthand for doing relational queries. The main idea would be to have joins, cartesian products, aggregations, operators, etc. that work by implicitly matching keys rather than by listing them. In fact, a lot of the APL operators would work if applied directly to tables, eg
orYou have to understand the frustration of running a search engine: you never know how good your result ranking is, because you never know which result the user clicked on, or whether the user simply gave up and went to another search engine.
The only way to track this behavior is to put up a redirect (or run an ISP - the free ones log just about everything).
Oops, another poster confirmed my suspicions: Rebecca Goldstein is the (ex?) wife of Shelly, the guy whose page I linked above.
Damned incestuous academic types.
Sorry I can't offer much insight, since I haven't read the book, but IIRC another review pegged this as David Bohm's story in disguise.
Bohm was a Princeton prof who revived de Broglie's Pilot Wave formulation, showed that it made sense, and then at some point was McCarthyized and ended up in Britain, blacklisted from US universities.
In any case, I like Bohm. He stacks up as follows:
Reality is:
deterministic, nonlocal, nonjumpy: Bohm, de Broglie
nondeterministic, local but jumpy: Bohr, Heisenberg, most basic physics books
deterministic, local: Einstein, Simpson (Homer)
nondeterministic, nonlocal, solipsistic: New Age physics-interpretation gurus
(sorry, it's Friday, hard to dredge this stuff up off the cuff)
Another random note: Most people don't know that de Broglie actually came up with his wave theory of matter to reconcile special relativity and the primitive quantum ideas of the time. The waves were a compensating factor for the slowdown of "internal vibrations" due to SR as v approaches c. So, in effect, QM is already a reconciliation with SR. However he eventually discarded his "pilot wave" theory, not because it didn't work, but because it didn't fit his assumptions about reality.
For more info on Bohmian Mechanics, here is a link to some current research.