The energy cost of building this seaborne city would be much greater than whatever savings it might obtain, whether built at sea or shipped there from a land base. How about the energy costs of moving people between this city and anyplace else, from which it would be remote?
Building on land isn't less energy efficient, it's more efficient. There's plenty of land near enough to oceans to take advantage of the ocean energy, without the higher costs of operating everything on the ocean. Any merit to these principles would be better applied to building a city on an island rather than a floating city from scratch.
This project is an obvious waste of time, money and energy. I smell a government grant sucked up by bankers and their grad student patsies.
No, you're a liar. When you repeat lies that so many others say are lies, but you don't bother to find out whether they're lies, you're a liar too. You don't have time to research the matter, but you have time to spin your lies on Slashdot? You're a liar.
At best you're a bullshitter: someone who doesn't care about truth or lies. Just spewing whatever you think helps you and hurts the ones you hate.
Why split hairs? The correct term is "Republican". Now tell me that you're "really a libertarian", or some other bullshit.
Media Matters received its first ever donation from Soros last month, after years of you Republicans lying, say he was financing it all along. Meanwhile, you Republicans have your fraud network financed by billionaires like the Koch brothers who also finance Republican campaigns, lately secretly through the Citizens United rules that dominated the election that just passed.
Of course it sounds "fair and balanced" to you, because it's an endless pile of Republican lies, just like the Fox "News" that uses that fraudulent slogan.
I don't see any evidence that this NLCP has ever called for investigating Halliburton. Even though the Bush/Cheney corruption with Halliburton was catastrophic for the country and totally obvious.
This NLCP attack is the kind of rightwing propaganda is the kind known as false equivalence. Rightwingers cook up some Democratic target to equate to some well known rightwing evildoer. The Clinton impeachment is a good example: Republicans live the legacy of Nixon resigning rather than face impeachment for very serious crimes, so they impeach a Democratic president the first chance they get. That way Republicans and the stupid people producing and consuming the mass media Republicans dominate can say "both sides do it", about either corruption/crimes or contrived impeachment. Even though there is no legitimate comparison between the two. For good measure, rightwingers will meet legitimate calls for Bush/Cheney's impeachment for lying us into the Iraq War by dismissing it falsely as equivalent to the Clinton impeachment. And it works: Clinton was impeached, Bush/Cheney were not.
And when Republicans impeach Obama on some nothing, they will claim falsely that it's equivalent to some Democratic action that is nothing of the kind.
Meanwhile, Bush peddles some fictional delusion book about his presidency, Halliburton continues to rape the world and the US (Gulf drilling catastrophe, anyone?) because it was never properly investigated, convicted and barred. Google continues to lead America's functional economy, and Obama continues to lead the US out of the hellhole Bush/Cheney and Halliburton dragged us into.
The attachment can include extra lenses. Or the capillary tubes can amplify the reaction to macroscopic scale. Indeed, the sample plates could turn color depending on what's inside, requiring only lorez cameras.
I don't see how that kind of mode change offers vendors or manufacturers the chance to change consumer expectations of rights in what they pay for. The mode change that is exploited for that is typically a hardware mode change, like mutually incompatible cartridges. Changes from keyboard to speech and pen/stylus aren't quite that kind of mode change.
In any case, we've gone pretty far afield in this discussion. It's been one of the longest I've had on Slashdot in perhaps several years that's remained a real conversation despite disagreements. Thanks for it.
Now that TV has gone digital, there's lots more spectrum even in the existing bands. As radio and DSP tech gets better, the same width radio bands carry more data, as subbands can be ever less far apart. Phased arrays don't need reserved spectrum at all, and there's all kinds of ways to shrink and eliminate towers, some of which I mentioned. And within 20 years, legacy equipment can be eliminated as quickly as it took to transition to digital TV. If it takes 25 years, that's not a meaningful difference in this discussion wherein I picked an arbitrary 20 years.
20 years is beyond a reasonable direct financial investment horizon if you're not market makers who can keep shaping those 20 years through continuing investment choices, as you and I are not. But it's still a timeframe I feel comfortable predicting in general outlines, because I've been doing it successfully in 5-10-15 year horizons since the 1970s. I don't know just who or how the speech/drawing inputs will be productized in 2030, which is required for financial investment. But I do have a good idea which technologies, products and techniques are consistent with the overall trend of telecom development over that duration.
In the 1980s, the NA videogame market was stagnant because the American console corps were complacent and paralyzed by cannibalistic management. At Atari in particular I know that cocaine in the marketing department and elsewhere was devastating all decisionmaking - and other competitors probably had equivalent problems I don't know about. Nintendo rebooted the industry, but only after trying and failing to sell something like NES or at least Donkey Kong to Atari, which notoriously turned up its complacent (and likely numb) nose, pressuring Nintendo to take the bigger risk and directly compete in NA. In any event, I don't really see what NES has to do with the ubiquity of mobile comms and the changing modes of communicating with computers that will bring.
People will use pen or fingertip to write as we did with ink when diagram or speech isn't acceptable. The big advance is intelligence in recognizing the variable input data, which will come over time, as it has come pretty far already. Parallel processing and DSP are just getting started on those problems, which have had more economical alternatives so far. But with pressures to dump the keyboard, as we already have with phones and tablets, the problem will be more pressing, and solutions will be developed faster and more completely.
Circuit design isn't like other programming, which is why it's not communicated to machines in a structured language like C. But for generations circuits were designed in schematics, not text. After diagram tools are popular among programmers higher-level than hardware designers, diagram tools will become popular among hardware designers, just as happened with text.
Diff and patch are harder in diagrams, but they're artifacts of text. What matters is change management. Topology is a more meaningful change to manage than text changes, which are arbitrary and harder for the human to process even when the differences are presented in text. Managing change and correlation among diagrams will be done in the terms of the diagrams. Those terms will allow more meaningful comparisons in terms of the actual work, instead of the more arbitrary correlations possible in text that describes the work much less manageably to start with.
Finding data, or rather symbols of data, from which to select is the most important problem now that data has become problematically abundant, while it was always problematically scarce before. Context, patterns, cross reference, and just accumulated wisdom about practice and meaning will give software the kind of ability to infer meaning from small cues just as humans do among ourselves. Likewise we'll become more forgiving of machine error as the interfaces to them become more expressive and personal. Which is the main reason humans communicate so "well" compared to HCI: we forgive each other's mistakes in ways we do not when we see them presented by a machine.
The desktop GUI hasn't changed fundamentally. But now that more people use phones than use computers, the limited hardware and different info and use cases it supports have non-desktops much more popular than are desktops across all users in the world. That change has already happened. The keyboard is an artifact of the desktop, as it was an artifact of the physical desktop that computers adopted as a metaphor. New UI metaphors show people and their relationships to each other, as well as other objects (like places) and their relationships (eg. maps). A desktop is a very limited work and social space, and people are already growing well past it. But it took 20 years for the desktop metaphor to be the main computer mode - even longer, as it wasn't until Windows 3.1 (1991) that most computer users even used a desktop most of the time, and really 1995 with Windows 95 that it was the overwhelming mode.
Sure, there are some East Asian languages that don't use ideographs. But Chinese/Japanese relationships of syllables to symbols are largely arbitrary, as the different spoken values in different Chinese languages (Han, Cantonese) each correspond to a single shared symbol set. Unlike Indoeuropean languages, in which text is always derived rather closely from the spoken version of the symbol. Billions of people are using symbols which are only incidentally spoken. The relevance to programming is that source code is rarely spoken aloud, and only incidentally.
Patents haven't prevented competition among game controllers, keyboards, mice, trackpads, trackballs, pointersticks, either between devices or between companies making the same device.
Yes. Because in 20 years those spectrum costs will be long ago repaid, and tower real estate will be long ago secured. Not to mention technology like phased arrays that eliminate the need for reserved spectrum and "ideal" tower locations or large towers. I also expect that flying radio platforms will replace most ground towers.
You are looking at all these bottlenecks as if they'd be pushed through tomorrow. 20 years is a long time, and the revenue to be gained is unparalleled. The real unknowns are business trends related to competition. Barring an extreme trend towards monopoly/cartel, rather than the opposite that is usual, market demand and tech development will make radio comms cheaper than wired comms on the fundamentals.
When speech is appropriate, people will speak to interfaces, whispering if necessary. When otherwise, people will gesture with a pen or finger.
The main change underlying these changes is that people will nearly always be selecting from existing content or "remixing" components. Mustering individual characters into words will rarely be necessary, and therefore rarely the mode used to communicate. Most text is already unnecessary, just a lazy or poorly designed UI that should already be selection from existing data already online. Inference from that data for new values will be even more possible as the critical mass just grows, as does templates for automated intelligence to use it.
Most people don't type 85WPM, or really "type" at all, but they use a keyboard instead of cursive. Speed is not the reason, even though some people do have an advantage. Some people will have speed advantage in the pointing/speaking modes, too, especially if it's advantageous in their lives/work.
Remember that we're talking about 20 years from now. There will still be programmers and other throwbacks with keyboards, just as there are still people who program on paper before typing. But for the overwhelming majority of the billions of people who will use computers all day long, they will be speaking and pointing. Certainly the billions of East Asians will not be typing QWERTY or even pinyin, and they will be the largest bloc of users. I used them as an example of people who do not form communication symbols from letters, or whose communication is directly mapped to speech at all. They use symbols and gestures, with contextual inference to add meaning to the symbols actually specified.
The hardware to do this won't be expensive in the timeframe I'm talking about. And indeed it doesn't matter which exact hardware implements this. I'm assuming that hardware will keep up with the requirements to let humans communicate in the most expressive but universally accessible methods available to us. That's speech and pen. The actual hardware will cope with that, in many ways that are not necessary to foresee in detail now in order to know the general way they will exist then.
Probably within 20 years mobile Internet will be like wired Internet: a single mobile device will access different networks operated by different operators. More like Europe's mobile networks rather than the current lockin. That allows competition, which allows lower prices. The actual costs of providing mobile Internet are lower than providing wired Internet, so those costs will go down to lower than wired broadband.
Or maybe not. If not, then whatever's keeping the mobile Internet safe from competition will probably apply to the wired Internet, too. In that environment where monopoly/cartel defines the economics, anything is possible, but since mobile Internet is cheaper to provide, it will still probably be cheaper than wired Internet.
In any case, mobile Internet will be pervasive, and therefore necessarily affordable. Of course, economics could wipe out all "discretionary" telecom by a middle class and all telecom by the poor, leaving only corporate users with either mobile Internet or any at all.
But there are very few plausible scenarios where the current price and quota structure lasts. It's just bad for business, since it keeps networks from exploiting the demand inherent in billions of humans. So that's not the way to bet.
Things are not at all clear in most procedural text software. They're more clear in diagrams.
The argument that "everything good is already how things are done" is defied by all kinds of developments. We've been trying to popularize tablets for over two decades, but only in the past year have they gained any traction. Xerox and SRI started showing people GUI/mouse/network/printer systems a decade and a half before they started to become popular. Now they're the vast majority of what people use, even if some tiny segment of users have a terminal window in one.
The way we program has changed, and the tools we use are not the best way to do it. It's inertia that has kept us from going to diagramming instead of procedures. Made mostly of people clinging to keyboards until death.
As for "that is your own problem", I will not be replying to any more of your posts, because you're a dick. Goodbye.
Saying computers will replace engineers is a lot riskier than saying that reused code will replace code from scratch. But indeed the fraction of people doing more than just entering data into computers (ie. "programming", even if it's more configuration than encoding logic) who do not write procedural code today overwhelms the fraction who do what programmers did 25 years ago. Almost all programming is at a higher level. Even hardcore programmers produce most of our code by selecting one or another package or library, or using a tool generating classes from data, rather than writing procedures. We're now writing a lot of XML that defines clients and servers in stubs, as well as UI.
There's been a lot of setbacks in going to graphs instead of statements. Microsoft's whole "Visual XYZ" coopted the momentum and then weighed it down with anything but graphs. Increased clock speed slowed down parallel hardware; simple client/server slowed down truly distributed processing in parallel networks. But most of those slowdowns have played out, while the hardware is now very parallel, as is the organization of people using and even producing the systems.
Even flowcharting has been too slow in use for defining the business that gets encoded. Businesspeople almost always want to describe their business in words, instead of diagrams, though the diagrams are more accurate and less prone to degrading when passed through design/production stages and cycles. These businesspeople all get a keyboard and a mouse to communicate, which are terrible for drawing diagrams. But now they're using touchscreens and pens. And there's now also years of libraries of business processes they can use as templates. We're just getting to a time where lots of business processes actually do start getting defined for the workers as well as the computers, and in diagrams as graphic artists are used more than "documentation writers". All of which streamlines the use of diagrams, which are better tools, produced from the beginning in machine readable format, which have further reach through the business and its engineering infrastructure.
The "change management" you described is more effective with graphs, since they're more universally understood throughout the organization. The automation necessary to support diagrams is a platform for distributing changes throughout the org and its products, instead of the illusion that "just change that function" means the changes don't need to be managed beyond the point they're originated. But until now, few people had the tools and skills to use graphs. Perhaps more essentially, they saw themselves as typing words rather than as drawing and changing diagrams. But mobile devices and more sophisticated input devices are changing that, so diagramming, shapes and simple references to existing content are becoming more natural for everyone.
It's the tools and skills that limit the ways people communicate. When all you have is a hammer, everything you do looks like a nail. When all you have is a keyboard and mouse (and desktop monitor), everything you do looks like a window of text. But now people have microphones, shape and gesture recognition in video, pens and pads for diagrams, and libraries of momentum in existing content to select from, instead of keyboards. That is where we're going.
Latex bodysuit and gas mask mainly mean abstinence, which is the only 100.0% safe sex. But marathon running - because it's not sex.
However, since there are other ways of contracting most STDs other than sex, even sexual abstinence isn't "100% safe", if you don't abstain from other possible causes.
It's all probabilities. Which for me and most people means some kind of sex is safe enough, and nothing is 100%.
I don't like how you're so stupid that you introduce your disagreement with an insult like "you're high".
The reason diagrammatic programming isn't popular is because we still use keyboards and mice, so diagramming is awkward. We also still allow way more original coding than is necessary or productive, instead of code reuse and references. Data streams are still the exception in the packaging of data, rather than data sets or data values, while classes/objects are still not really models of the work but rather packages of data and operations that more reflect the code production process than the objects/relations natural to the work.
But both the input hardware and the code to refer to are evolving far enough that they're starting to retain "from scratch" procedural practice more as convention than as convenience. 20 years from now, especially with so many East Asian programmers and stakeholders already familiar with symbolic rather than alphabetic representation, programmers will pattern computer processes by indicating relations to existing data paths and processing stages much more than stating in words that can be spoken and spelled how to operate on new collections of bits.
It's still early in HCI development, but 20 years from now won't be. All the original keyboard programmers will be dead, teenage programmers will be typical instead of exceptional, and most professionals worldwide will be at least configuring computers, always using them for practically all communications, instead of the pen/paper that still dominates in places like the medical and field service industries.
Getting one of these heaters into the lap of every child in the world might just be our best chance to keep humans for further overpopulating the planet to catastrophe.
Or maybe that's the Internet's plan to extinct us and take over. Phase I, replacing our reproductive drive with porn, is already wildly successful.
People who pee on their phones are less likely to get any sexually transmitted disease, because they're less likely to get any sex (with another person, anyway).
However, those phone peeing people who do get sex are more likely to catch something dirty, given the kind of people who will have sex with them.
The use stats of a device like this could tell us quite a lot about human nature.
How about a little rig with capillary tubes etched throughout a couple of thin plastic plates that you touch edgewise to some urine or blood, that pulls the fluid through, then snaps into a little frame attached to the phone's camera lens. All calibrated to give image data to a server that looks for interactions of disease causes/products with the sizes, shapes and materials in the tubes. Then sends results back to the phone. The little rig should be small and cheap enough to dispense in nightclub bathrooms or drugstores, neater than a pregnancy test, and without leaving any analysis up to the user's eyesight, manual dexterity or intelligence.
The people we most want getting prompt STD infection results are the ones who already aren't competent to keep safe by practicing safe sex. And other infectious diseases are just a little further back in the "evolution safety skills" stack. "Foolproof" is the #1 design objective, because fools have a higher rate of being the most important user segment.
Slashdot Mirror
The energy cost of building this seaborne city would be much greater than whatever savings it might obtain, whether built at sea or shipped there from a land base. How about the energy costs of moving people between this city and anyplace else, from which it would be remote?
Building on land isn't less energy efficient, it's more efficient. There's plenty of land near enough to oceans to take advantage of the ocean energy, without the higher costs of operating everything on the ocean. Any merit to these principles would be better applied to building a city on an island rather than a floating city from scratch.
This project is an obvious waste of time, money and energy. I smell a government grant sucked up by bankers and their grad student patsies.
No, you're a liar. When you repeat lies that so many others say are lies, but you don't bother to find out whether they're lies, you're a liar too. You don't have time to research the matter, but you have time to spin your lies on Slashdot? You're a liar.
At best you're a bullshitter: someone who doesn't care about truth or lies. Just spewing whatever you think helps you and hurts the ones you hate.
Why split hairs? The correct term is "Republican". Now tell me that you're "really a libertarian", or some other bullshit.
Working. And taking time off while Slashdot's signal:noise ratio went through the floor. Now it's back on the floor, and worth wasting some time here.
No, you lie.
Media Matters received its first ever donation from Soros last month, after years of you Republicans lying, say he was financing it all along. Meanwhile, you Republicans have your fraud network financed by billionaires like the Koch brothers who also finance Republican campaigns, lately secretly through the Citizens United rules that dominated the election that just passed.
Of course it sounds "fair and balanced" to you, because it's an endless pile of Republican lies, just like the Fox "News" that uses that fraudulent slogan.
No, they're not. Except insofar as "false" is a wrong version of "true".
I don't see any evidence that this NLCP has ever called for investigating Halliburton. Even though the Bush/Cheney corruption with Halliburton was catastrophic for the country and totally obvious.
This NLCP attack is the kind of rightwing propaganda is the kind known as false equivalence. Rightwingers cook up some Democratic target to equate to some well known rightwing evildoer. The Clinton impeachment is a good example: Republicans live the legacy of Nixon resigning rather than face impeachment for very serious crimes, so they impeach a Democratic president the first chance they get. That way Republicans and the stupid people producing and consuming the mass media Republicans dominate can say "both sides do it", about either corruption/crimes or contrived impeachment. Even though there is no legitimate comparison between the two. For good measure, rightwingers will meet legitimate calls for Bush/Cheney's impeachment for lying us into the Iraq War by dismissing it falsely as equivalent to the Clinton impeachment. And it works: Clinton was impeached, Bush/Cheney were not.
And when Republicans impeach Obama on some nothing, they will claim falsely that it's equivalent to some Democratic action that is nothing of the kind.
Meanwhile, Bush peddles some fictional delusion book about his presidency, Halliburton continues to rape the world and the US (Gulf drilling catastrophe, anyone?) because it was never properly investigated, convicted and barred. Google continues to lead America's functional economy, and Obama continues to lead the US out of the hellhole Bush/Cheney and Halliburton dragged us into.
Your Republican America at work.
The attachment can include extra lenses. Or the capillary tubes can amplify the reaction to macroscopic scale. Indeed, the sample plates could turn color depending on what's inside, requiring only lorez cameras.
Kill yourself.
I don't see how that kind of mode change offers vendors or manufacturers the chance to change consumer expectations of rights in what they pay for. The mode change that is exploited for that is typically a hardware mode change, like mutually incompatible cartridges. Changes from keyboard to speech and pen/stylus aren't quite that kind of mode change.
In any case, we've gone pretty far afield in this discussion. It's been one of the longest I've had on Slashdot in perhaps several years that's remained a real conversation despite disagreements. Thanks for it.
Now that TV has gone digital, there's lots more spectrum even in the existing bands. As radio and DSP tech gets better, the same width radio bands carry more data, as subbands can be ever less far apart. Phased arrays don't need reserved spectrum at all, and there's all kinds of ways to shrink and eliminate towers, some of which I mentioned. And within 20 years, legacy equipment can be eliminated as quickly as it took to transition to digital TV. If it takes 25 years, that's not a meaningful difference in this discussion wherein I picked an arbitrary 20 years.
20 years is beyond a reasonable direct financial investment horizon if you're not market makers who can keep shaping those 20 years through continuing investment choices, as you and I are not. But it's still a timeframe I feel comfortable predicting in general outlines, because I've been doing it successfully in 5-10-15 year horizons since the 1970s. I don't know just who or how the speech/drawing inputs will be productized in 2030, which is required for financial investment. But I do have a good idea which technologies, products and techniques are consistent with the overall trend of telecom development over that duration.
In the 1980s, the NA videogame market was stagnant because the American console corps were complacent and paralyzed by cannibalistic management. At Atari in particular I know that cocaine in the marketing department and elsewhere was devastating all decisionmaking - and other competitors probably had equivalent problems I don't know about. Nintendo rebooted the industry, but only after trying and failing to sell something like NES or at least Donkey Kong to Atari, which notoriously turned up its complacent (and likely numb) nose, pressuring Nintendo to take the bigger risk and directly compete in NA. In any event, I don't really see what NES has to do with the ubiquity of mobile comms and the changing modes of communicating with computers that will bring.
People will use pen or fingertip to write as we did with ink when diagram or speech isn't acceptable. The big advance is intelligence in recognizing the variable input data, which will come over time, as it has come pretty far already. Parallel processing and DSP are just getting started on those problems, which have had more economical alternatives so far. But with pressures to dump the keyboard, as we already have with phones and tablets, the problem will be more pressing, and solutions will be developed faster and more completely.
Circuit design isn't like other programming, which is why it's not communicated to machines in a structured language like C. But for generations circuits were designed in schematics, not text. After diagram tools are popular among programmers higher-level than hardware designers, diagram tools will become popular among hardware designers, just as happened with text.
Diff and patch are harder in diagrams, but they're artifacts of text. What matters is change management. Topology is a more meaningful change to manage than text changes, which are arbitrary and harder for the human to process even when the differences are presented in text. Managing change and correlation among diagrams will be done in the terms of the diagrams. Those terms will allow more meaningful comparisons in terms of the actual work, instead of the more arbitrary correlations possible in text that describes the work much less manageably to start with.
Finding data, or rather symbols of data, from which to select is the most important problem now that data has become problematically abundant, while it was always problematically scarce before. Context, patterns, cross reference, and just accumulated wisdom about practice and meaning will give software the kind of ability to infer meaning from small cues just as humans do among ourselves. Likewise we'll become more forgiving of machine error as the interfaces to them become more expressive and personal. Which is the main reason humans communicate so "well" compared to HCI: we forgive each other's mistakes in ways we do not when we see them presented by a machine.
The desktop GUI hasn't changed fundamentally. But now that more people use phones than use computers, the limited hardware and different info and use cases it supports have non-desktops much more popular than are desktops across all users in the world. That change has already happened. The keyboard is an artifact of the desktop, as it was an artifact of the physical desktop that computers adopted as a metaphor. New UI metaphors show people and their relationships to each other, as well as other objects (like places) and their relationships (eg. maps). A desktop is a very limited work and social space, and people are already growing well past it. But it took 20 years for the desktop metaphor to be the main computer mode - even longer, as it wasn't until Windows 3.1 (1991) that most computer users even used a desktop most of the time, and really 1995 with Windows 95 that it was the overwhelming mode.
Sure, there are some East Asian languages that don't use ideographs. But Chinese/Japanese relationships of syllables to symbols are largely arbitrary, as the different spoken values in different Chinese languages (Han, Cantonese) each correspond to a single shared symbol set. Unlike Indoeuropean languages, in which text is always derived rather closely from the spoken version of the symbol. Billions of people are using symbols which are only incidentally spoken. The relevance to programming is that source code is rarely spoken aloud, and only incidentally.
Patents haven't prevented competition among game controllers, keyboards, mice, trackpads, trackballs, pointersticks, either between devices or between companies making the same device.
Yes. Because in 20 years those spectrum costs will be long ago repaid, and tower real estate will be long ago secured. Not to mention technology like phased arrays that eliminate the need for reserved spectrum and "ideal" tower locations or large towers. I also expect that flying radio platforms will replace most ground towers.
You are looking at all these bottlenecks as if they'd be pushed through tomorrow. 20 years is a long time, and the revenue to be gained is unparalleled. The real unknowns are business trends related to competition. Barring an extreme trend towards monopoly/cartel, rather than the opposite that is usual, market demand and tech development will make radio comms cheaper than wired comms on the fundamentals.
When speech is appropriate, people will speak to interfaces, whispering if necessary. When otherwise, people will gesture with a pen or finger.
The main change underlying these changes is that people will nearly always be selecting from existing content or "remixing" components. Mustering individual characters into words will rarely be necessary, and therefore rarely the mode used to communicate. Most text is already unnecessary, just a lazy or poorly designed UI that should already be selection from existing data already online. Inference from that data for new values will be even more possible as the critical mass just grows, as does templates for automated intelligence to use it.
Most people don't type 85WPM, or really "type" at all, but they use a keyboard instead of cursive. Speed is not the reason, even though some people do have an advantage. Some people will have speed advantage in the pointing/speaking modes, too, especially if it's advantageous in their lives/work.
Remember that we're talking about 20 years from now. There will still be programmers and other throwbacks with keyboards, just as there are still people who program on paper before typing. But for the overwhelming majority of the billions of people who will use computers all day long, they will be speaking and pointing. Certainly the billions of East Asians will not be typing QWERTY or even pinyin, and they will be the largest bloc of users. I used them as an example of people who do not form communication symbols from letters, or whose communication is directly mapped to speech at all. They use symbols and gestures, with contextual inference to add meaning to the symbols actually specified.
The hardware to do this won't be expensive in the timeframe I'm talking about. And indeed it doesn't matter which exact hardware implements this. I'm assuming that hardware will keep up with the requirements to let humans communicate in the most expressive but universally accessible methods available to us. That's speech and pen. The actual hardware will cope with that, in many ways that are not necessary to foresee in detail now in order to know the general way they will exist then.
But it's still where the growth is.
Probably within 20 years mobile Internet will be like wired Internet: a single mobile device will access different networks operated by different operators. More like Europe's mobile networks rather than the current lockin. That allows competition, which allows lower prices. The actual costs of providing mobile Internet are lower than providing wired Internet, so those costs will go down to lower than wired broadband.
Or maybe not. If not, then whatever's keeping the mobile Internet safe from competition will probably apply to the wired Internet, too. In that environment where monopoly/cartel defines the economics, anything is possible, but since mobile Internet is cheaper to provide, it will still probably be cheaper than wired Internet.
In any case, mobile Internet will be pervasive, and therefore necessarily affordable. Of course, economics could wipe out all "discretionary" telecom by a middle class and all telecom by the poor, leaving only corporate users with either mobile Internet or any at all.
But there are very few plausible scenarios where the current price and quota structure lasts. It's just bad for business, since it keeps networks from exploiting the demand inherent in billions of humans. So that's not the way to bet.
Things are not at all clear in most procedural text software. They're more clear in diagrams.
The argument that "everything good is already how things are done" is defied by all kinds of developments. We've been trying to popularize tablets for over two decades, but only in the past year have they gained any traction. Xerox and SRI started showing people GUI/mouse/network/printer systems a decade and a half before they started to become popular. Now they're the vast majority of what people use, even if some tiny segment of users have a terminal window in one.
The way we program has changed, and the tools we use are not the best way to do it. It's inertia that has kept us from going to diagramming instead of procedures. Made mostly of people clinging to keyboards until death.
As for "that is your own problem", I will not be replying to any more of your posts, because you're a dick. Goodbye.
Saying computers will replace engineers is a lot riskier than saying that reused code will replace code from scratch. But indeed the fraction of people doing more than just entering data into computers (ie. "programming", even if it's more configuration than encoding logic) who do not write procedural code today overwhelms the fraction who do what programmers did 25 years ago. Almost all programming is at a higher level. Even hardcore programmers produce most of our code by selecting one or another package or library, or using a tool generating classes from data, rather than writing procedures. We're now writing a lot of XML that defines clients and servers in stubs, as well as UI.
There's been a lot of setbacks in going to graphs instead of statements. Microsoft's whole "Visual XYZ" coopted the momentum and then weighed it down with anything but graphs. Increased clock speed slowed down parallel hardware; simple client/server slowed down truly distributed processing in parallel networks. But most of those slowdowns have played out, while the hardware is now very parallel, as is the organization of people using and even producing the systems.
Even flowcharting has been too slow in use for defining the business that gets encoded. Businesspeople almost always want to describe their business in words, instead of diagrams, though the diagrams are more accurate and less prone to degrading when passed through design/production stages and cycles. These businesspeople all get a keyboard and a mouse to communicate, which are terrible for drawing diagrams. But now they're using touchscreens and pens. And there's now also years of libraries of business processes they can use as templates. We're just getting to a time where lots of business processes actually do start getting defined for the workers as well as the computers, and in diagrams as graphic artists are used more than "documentation writers". All of which streamlines the use of diagrams, which are better tools, produced from the beginning in machine readable format, which have further reach through the business and its engineering infrastructure.
The "change management" you described is more effective with graphs, since they're more universally understood throughout the organization. The automation necessary to support diagrams is a platform for distributing changes throughout the org and its products, instead of the illusion that "just change that function" means the changes don't need to be managed beyond the point they're originated. But until now, few people had the tools and skills to use graphs. Perhaps more essentially, they saw themselves as typing words rather than as drawing and changing diagrams. But mobile devices and more sophisticated input devices are changing that, so diagramming, shapes and simple references to existing content are becoming more natural for everyone.
It's the tools and skills that limit the ways people communicate. When all you have is a hammer, everything you do looks like a nail. When all you have is a keyboard and mouse (and desktop monitor), everything you do looks like a window of text. But now people have microphones, shape and gesture recognition in video, pens and pads for diagrams, and libraries of momentum in existing content to select from, instead of keyboards. That is where we're going.
No, you're a fool.
Latex bodysuit and gas mask mainly mean abstinence, which is the only 100.0% safe sex. But marathon running - because it's not sex.
However, since there are other ways of contracting most STDs other than sex, even sexual abstinence isn't "100% safe", if you don't abstain from other possible causes.
It's all probabilities. Which for me and most people means some kind of sex is safe enough, and nothing is 100%.
How about an actual argument, instead of just an anonymous disagreement?
Thanks: let me know how foolproof the .sig is ;).
Anonymous crackhead Coward, you don't need a keyboard to input "LOL". You're as smug as you are stupid.
I don't like how you're so stupid that you introduce your disagreement with an insult like "you're high".
The reason diagrammatic programming isn't popular is because we still use keyboards and mice, so diagramming is awkward. We also still allow way more original coding than is necessary or productive, instead of code reuse and references. Data streams are still the exception in the packaging of data, rather than data sets or data values, while classes/objects are still not really models of the work but rather packages of data and operations that more reflect the code production process than the objects/relations natural to the work.
But both the input hardware and the code to refer to are evolving far enough that they're starting to retain "from scratch" procedural practice more as convention than as convenience. 20 years from now, especially with so many East Asian programmers and stakeholders already familiar with symbolic rather than alphabetic representation, programmers will pattern computer processes by indicating relations to existing data paths and processing stages much more than stating in words that can be spoken and spelled how to operate on new collections of bits.
It's still early in HCI development, but 20 years from now won't be. All the original keyboard programmers will be dead, teenage programmers will be typical instead of exceptional, and most professionals worldwide will be at least configuring computers, always using them for practically all communications, instead of the pen/paper that still dominates in places like the medical and field service industries.
Getting one of these heaters into the lap of every child in the world might just be our best chance to keep humans for further overpopulating the planet to catastrophe.
Or maybe that's the Internet's plan to extinct us and take over. Phase I, replacing our reproductive drive with porn, is already wildly successful.
People who pee on their phones are less likely to get any sexually transmitted disease, because they're less likely to get any sex (with another person, anyway).
However, those phone peeing people who do get sex are more likely to catch something dirty, given the kind of people who will have sex with them.
The use stats of a device like this could tell us quite a lot about human nature.
How about a little rig with capillary tubes etched throughout a couple of thin plastic plates that you touch edgewise to some urine or blood, that pulls the fluid through, then snaps into a little frame attached to the phone's camera lens. All calibrated to give image data to a server that looks for interactions of disease causes/products with the sizes, shapes and materials in the tubes. Then sends results back to the phone. The little rig should be small and cheap enough to dispense in nightclub bathrooms or drugstores, neater than a pregnancy test, and without leaving any analysis up to the user's eyesight, manual dexterity or intelligence.
The people we most want getting prompt STD infection results are the ones who already aren't competent to keep safe by practicing safe sex. And other infectious diseases are just a little further back in the "evolution safety skills" stack. "Foolproof" is the #1 design objective, because fools have a higher rate of being the most important user segment.