I forgot to add: within five minutes of installing 11.10 the first time and the second time, I had borked it by changing some settings using the supplied applet. There was no panel and no menu bar: just the bare desktop background. The system didn't respond to key presses or mouse clicks. Rebooting did not change matters.
Again, not a problem for people with a copy of another linux distro to use. But there's no way it should be possible to do what I did. (The experience soured me on Ubuntu to the point that even if I could remember how I did it, I wouldn' bother reporting the bug.)
What's wrong with Unity? Its usability and its fragility.
The problem is having the menu bar at the top of the screen together with 'focus follows mouse'.
If you have overlapping windows, or one below another, and use the mouse to go from a lower window up to its menus, as the mouse passes over the higher-up window the menu bar changes to hold the menus for that app.
And you don't see that, because the menus aren't shown until the mouse actually gets to the menu bar.
Having menus at the top of the screen instead of within the app window also significantly increases the distance the mouse has to travel to get to them.
Having the menus not shown until the mouse gets to them means you can't aim for the menu you want. Getting to the menu is a three-step process: move the mouse pointer to the top of the screen, read the menus, then adjust horizontally to the menu you want.
Things not visible; things changing unexpectedly; making the user work harder. All in all, usability failure.
(And don't tell me to use F10. Ubuntu doesn't give me the option to configure my Apple compact USB keyboard to prefer F1..F12 to multimedia functions, so that's fn + F10. And f'n everything else; control-alt-T doesn't work, for instance. I can fix these things, but Ubuntu's target user can't.)
I thought I'd give Fedora 15 a try. I couldn't get it to recognise my software RAID5 (on which I keep/home,/srv, and a few other things). The base drive partitions just weren't listed in/proc/partitions; nor was the md device.
OK - the partitions are of type 0xda; I set up the RAID a long time ago, and that was the advice, then. But so what? Why should that keep them out of/proc/partitions?
Shouldn't be too hard to fix. But it was. Fedora's documentation is extremely poorly organised, and the page design is not very usable. Google wasn't any help; I still had to spend far too long trying to navigate my way around weirdly organised documentation. It felt like the Fedora project didn't actually want users.
I went back to Arch.
Arch and Debian are the only two distributions I've used that have anything approaching conceptual unity. Slackware and Gentoo may have it too; I don't know. One day I'll have a look, perhaps.
Arch is by far the best in terms of conceptual unity. Even Debian is only moderately good in this respect. The others are just fruit salads. I don't have the inclination to learn seventeen different ways of implementing configuration files, or to google to find out that something that should be in/etc is, in this distro, in/usr/share, or any of the other stupid quirks.
Fedora was good for one thing: it reminded me again why I've been using Arch for the last five years.
In 1900 in London, England -- the heart of the industrial revolution at its culmination -- about 25 percent of the population had no contact with the formal economy. A quarter of households weren't even unemployed.
"Disruption"? Malnourishment, chronic disease, illiteracy, crime. The grandparents and great-grandparents of these people were skilled.
The way it works is that one or more generations suffer "disruption", and then new products and industries get developed and employ the descendants of the disrupted....perhaps. If political elites see it as in their interests to provide services and income to the underclasses.
If Bettencourt, Lobo, et al. (2007) are correct, cycles of innovation such as the textile revolution will have to repeat at an ever-increasing tempo if we are to avoid collapse.
Oh, and that food surplus? It was produced by doubling the arable land of the world, by means of wiping out the native Americans. Britain in the 1860s could only produce enough food for 8 million of its 21 million inhabitants. (Gregory Clark, 2002.)
First agriculture, then manufacturing, then services.
In 1900, agriculture employed 55% of the civilian labor force. Now, about 2%, depending how you count seasonal labor. In 1970 manufacturing was about 35% (IIRC) of US employment. Now, 14% and falling. In 2011, services employ over 70% of the labor force. In 2041,..?
In services the big occupation is "cashier", with "retail supervisor" next (again, IIRC. Check the Bureau of Labor Statistics website if interested.) A comment up-thread talked about bus- and taxi-drivers, but truck-driving employs many more people - it's in the top ten of occupations.
It's not too hard to imagine these three occupations being automated. The first supermarket chain that figures out how to do for food what Amazon has done for books is going to make a killing. (Speaking of Amazon, automating its warehouses can't be more than a decade off.) And people are buying more and more take-out food, clothes and accessories on-line.
What can displaced cashiers and truck drivers do for income? That's where I get stuck.
The assumptions made by the people that assert "there will be jobs" are these:
1. Robot vision and statistical machine learning are dead ends and won't develop any further. 2. The space of human desires is unbounded. 3. Price elasticity of demand in aggregate is greater than one. 4. We will reintroduce checks to the tendency of markets to concentrate wealth and income.
Assumptions 2, 3, and 4 are of unknown validity. Assumption 1 is almost certainly wrong.
Little-known fact: the US has around 250,000 producing wells. However, the vast majority produce less than 10 barrels per day. IIRC the total number of wells drilled is over 400,000.
With 2,000,000 wells each producing 10 barrels, the USA will be set - provided that electricity stays cheap enough to operate all those well pumps. At present there are 1200 drilling rigs which can each drill two wells per year. So assuming no dry holes and indefinite well life, in 729 years: energy security!
Static analyses are great fun, but they are misleading.
PV cells have been decreasing in cost at the rate of 22% per doubling of production capacity for three decades now. There are good reasons to think that this trend will continue. Since PV now provides less than 0.03% of global energy, there's plenty of room for ten more doublings. That gets us down to under ten percent of current PV costs.
Balance-of-system costs (inverters, support structures, installation costs, and especially permits/approvals) have decreased more slowly than PV costs in the past, primarily because they used to be negligible. Now they are about the same size as PV cost, and lo and behold, people are starting to work on getting them down.
Now, details.
"1 ton of coal costs $36 = $0.006 per KWH"
So why is it that coal-fired power stations don't charge any less than $0.04 per kWh? Highway robbery!
You are -- or rather Green Econ is, please put quote markers on your quotes, and quote them properly -- comparing the cost of the coal to the capital cost of a PV plant, not the cost of its fuel, sunlight. The correct comparison is the capital costs of coal mines, railroads, and power stations versus the capital costs of PV installations. Capital cost is why coal-fired power stations charge 0.04, and they couldn't charge much less if the coal was free.
Green Econ is a shill for the coal industry, but only fools the uncritical. Good critical thinking practice for you!
"Historical records have had the average temperature warmer than now"
That may be true in another world of the many-worlds multiverse. Not this one.
Not globally, not even for the Northern Hemisphere, not for any climatically meaningful interpretation of "now".
Really, people, this is not hard. Google for Spencer Weart, read his website, then google Skeptical Science and read John Cook's web site.
"the ice caps . . . just three years ago, people were claiming they'd be gone in six [years]..."
Aha. I see the problem: reading comprehension. It was not ice caps but Arctic sea ice that was exercising the imaginations of bloggers. Cryosphere researchers expect the ice caps to last thousands of years -- tens of thousands of years, in the case of the Antarctic ice cap.
Humans need energy too, you know. Yes, the chain of work from sun->plants->other stuff->humans->human labor is highly energy efficient, but do you really think it's impossible to approach that level of efficiency artificially?
Kidding, right? Sun->plants: about 0.5% in-the-field efficiency for the harvested bits, at best. Plants->other stuff: about 70% efficient for harvest losses, transport and storage losses, and pre-sale losses. Oh, you wanted milk, eggs or meat? 10% efficiency on average. Other stuff->humans: about 60% to 80% efficient. Humans->human labor: maximum eight-hour rate, about 10% efficient. Overall: 0.05 * 0.7 *0.5 * 0.8 *0.1 = 0.14%. One part in one-thousandth.
Yes, I think we might be able to do as well as that with photovoltaic panels and electric motors.
If you're past point 1, then at the risk of starting a flame war:-
For a compute-bound problem that *can't go on GPUs* and needs fast turnaround, you'd be silly to use anything other than Sandy Bridge at the moment. Core i7 2600 if you can tolerate consumer grade (keep a spare) or the equivalent Xeon, or better. (See, I told you, flames! Look, guys, I like AMD, I want them to succeed, but benchmarks are benchmarks.) For the combination of raw speed, FLOPS/watt, and minimum idle power, Sandy Bridge beats everything else at present. If you're paying for the power, you'll maximise your total computation per pound with SNB. And be much more likely to have your overnight jobs finished and waiting for you when you come in in the morning.
Regarding second-hand stuff, even if you can get it for free, if it's older/smaller than Core 2 Quad Q6600 or the equivalent Xeon, pass on it. There aren't enough cpu instructions executed per second or per kilowatt-hour in anything older. Some of the older cases can be re-used, though.
Spend the money on a programmer to parallelize the algorithm on standard CPUs, and put it out on BOINC.
No, DON'T use BOINC. It will take six months to set up the project and publicise it, and a further year to learn how to deal with the volunteers so as to get work done reasonably reliably. Half or more of the work you send out will never come back. Newbs will bother you with questions about BOINC and the sixteen other projects they're 'donating time' to, not your project. (Virus scanners cause people grief.) RAC hunters will badger you incessantly about minor discrepancies in points awarded.
Running a project through BOINC and getting value out of it means committing a *lot* of time and effort just on the PR side of the project-- setting up and maintaining a website and the Boinc infrastructure, and doing regular news updates, answering emails, getting rid of spammers, and so on.
Unless, of course, you don't get any volunteers.
Unless your work is very interesting to the general public and you can use a big PR machine, trying to use BOINC is a good way to waste a year and achieve nothing.
Better hope your manager doesn't read Slashdot. She'll be asking herself, "what use have I got for someone who takes weeks to realise he needs to use his editor's 'show white space characters' command when dealing with a white space sensitive language?"
In December 2010, Molycorp announced that it secured all the environmental permits needed to begin construction of a new ore processing plant at the mine; construction will begin in January 2011, and is expected to be completed by the end of 2012.[17]
I have a Core i7 2600 and a Velociraptor, and LXDE is noticeably faster than Xfce, let alone Gnome 2 or KDE. So much so that I always go back to it, despite its comparative lack of features, and I'm no hyperactive multitasker.
Speed has been the number 1 user interface problem ever since the days of TTYs. And it probably always will be. I'm surprised someone like Torvalds isn't already using LXDE.
you can buy a near infinite amount of services and be environmentally sustainable.
Only if the services are completely intangible, for example buying a Vuitton branded plastic grocery bag for $1M rather than an unbranded one for $0.05. Nearly all services consume energy and other resources in the act, and competition keeps their price at the marginal cost of supply. We can only increase the amount of energy consumed about 100 times before heat pollution starts to get serious, and prices can't rise arbitrarily.
I keep wondering too. Haven't heard a satisfactory answer in five years of asking.
Some of the more sensible things that have been suggested: arts and entertainment - everybody can make money by being famous (yeah, right); sports - a subcategory of entertainment; personal services: be a lifestyle coach or language tutor or personal shopper;... or... uh, that's about it, really.
A job that I hope returns in large numbers: union delegate. But I expect to wait at least twenty years.
Specialization is a pretty basic to macroeconomics.
Yes! That explains why the Germans make cars and sell them to the French, and in return the French make... cars, and... and sell them to the, the... the Germans...
Wait, what?
I think you mean, specialization is fundamental to increaasing productivity at the firm and industry level. That is, in microeconomics.
The jobs we outsource are jobs we don't actually want because they don't pay well enough and are pretty unskilled.
Of course. That explains why since the recession bottomed out, nearly all the growth in employment in the USA has been in jobs that pay less than the median wage.... Oh, wait.
A $20K car from 1980 is certainly higher quality than a $20k car from today, too.
Actually, no. A vanilla Honda Civic of today has vastly better suspension, brakes, steering, and headlamps; it has side-intrusion beams and airbags; it's much slower to rust or fade; and it's much more reliable than a $20,000 luxury sports car of 1981. Its acceleration isn't quite as good, but it's not as far back as you might think.
This is leaving aside frills like electric windows, air conditioning, sound damping (so that you can actually hear the stereo), and remote central locking, some of which the older car may have had, briefly. The Civic won't have leather seats, a chrome-plated ashtray or a walnut knob on the stick shift, but I know which I'd prefer to use daily.
Extrapolating from cars, we should have much better value from sound systems than we do. The fact that we don't have any improvement is because most people don't care.
IMHO, there's nothing that sounds quite as nice as wood for a speaker enclosure
Concrete is better -- it has very little in the way of resonances at the energies and frequencies experienced in speakers. Your speakers sound more like the design calcs say they should. But we're way into DIY territory here, lining the insides of our speakers with an inch of concrete all 'round.
You're right that the speakers are far and away the most important determinant of the sound. The rule of thumb is, half your budget on the speakers, and the rest for everything else.
The young, female human ear can hear up to 22 kHz. (Off topic: why can't people use SI prefixes properly, goddammit? "Kilo" is "k", not "K". Mega is M, not m.)
The male ear over 20 can't hear much more than 19 kHz, considerably lower if it was brought up in an urban environment. And it's all downhill from there. By 50, you're lucky if you can hear much over 9 kHz.
Shannon’s Law says that a digital signal is indistinguishable from an analog signal at double the frequency of the analog signal
No, it doesn't.
Shannon's law says you can recover a sine wave from a sample at double the wave's frequency. Music and speech are not sine waves.
Hint: in professional sound recording, sampling is done at 192 kHz because sound engineers can hear a difference between that and 96 kHz, let alone {shudder} 44.1 kHz. They don't do it for fun: the files are double the bulk, so the engineers have to truck around more storage and wait longer for files to transfer. Time is money. But they use 192 kHz nonetheless.
Slashdot Mirror
I forgot to add: within five minutes of installing 11.10 the first time and the second time, I had borked it by changing some settings using the supplied applet. There was no panel and no menu bar: just the bare desktop background. The system didn't respond to key presses or mouse clicks. Rebooting did not change matters.
Again, not a problem for people with a copy of another linux distro to use. But there's no way it should be possible to do what I did. (The experience soured me on Ubuntu to the point that even if I could remember how I did it, I wouldn' bother reporting the bug.)
What's wrong with Unity? Its usability and its fragility.
The problem is having the menu bar at the top of the screen together with 'focus follows mouse'.
If you have overlapping windows, or one below another, and use the mouse to go from a lower window up to its menus, as the mouse passes over the higher-up window the menu bar changes to hold the menus for that app.
And you don't see that, because the menus aren't shown until the mouse actually gets to the menu bar.
Having menus at the top of the screen instead of within the app window also significantly increases the distance the mouse has to travel to get to them.
Having the menus not shown until the mouse gets to them means you can't aim for the menu you want. Getting to the menu is a three-step process: move the mouse pointer to the top of the screen, read the menus, then adjust horizontally to the menu you want.
Things not visible; things changing unexpectedly; making the user work harder. All in all, usability failure.
(And don't tell me to use F10. Ubuntu doesn't give me the option to configure my Apple compact USB keyboard to prefer F1..F12 to multimedia functions, so that's fn + F10. And f'n everything else; control-alt-T doesn't work, for instance. I can fix these things, but Ubuntu's target user can't.)
Yeah, how about it?
I thought I'd give Fedora 15 a try. I couldn't get it to recognise my software RAID5 (on which I keep /home, /srv, and a few other things). The base drive partitions just weren't listed in /proc/partitions; nor was the md device.
OK - the partitions are of type 0xda; I set up the RAID a long time ago, and that was the advice, then. But so what? Why should that keep them out of /proc/partitions?
Shouldn't be too hard to fix. But it was. Fedora's documentation is extremely poorly organised, and the page design is not very usable. Google wasn't any help; I still had to spend far too long trying to navigate my way around weirdly organised documentation. It felt like the Fedora project didn't actually want users.
I went back to Arch.
Arch and Debian are the only two distributions I've used that have anything approaching conceptual unity. Slackware and Gentoo may have it too; I don't know. One day I'll have a look, perhaps.
Arch is by far the best in terms of conceptual unity. Even Debian is only moderately good in this respect. The others are just fruit salads. I don't have the inclination to learn seventeen different ways of implementing configuration files, or to google to find out that something that should be in /etc is, in this distro, in /usr/share, or any of the other stupid quirks.
Fedora was good for one thing: it reminded me again why I've been using Arch for the last five years.
In 1900 in London, England -- the heart of the industrial revolution at its culmination -- about 25 percent of the population had no contact with the formal economy. A quarter of households weren't even unemployed.
"Disruption"? Malnourishment, chronic disease, illiteracy, crime. The grandparents and great-grandparents of these people were skilled.
The way it works is that one or more generations suffer "disruption", and then new products and industries get developed and employ the descendants of the disrupted....perhaps. If political elites see it as in their interests to provide services and income to the underclasses.
If Bettencourt, Lobo, et al. (2007) are correct, cycles of innovation such as the textile revolution will have to repeat at an ever-increasing tempo if we are to avoid collapse.
Oh, and that food surplus? It was produced by doubling the arable land of the world, by means of wiping out the native Americans. Britain in the 1860s could only produce enough food for 8 million of its 21 million inhabitants. (Gregory Clark, 2002.)
Not the economic system, the political system. The willingness to redistribute income is a political decision.
Economics has something to say about methods and their consequences only.
First agriculture, then manufacturing, then services.
In 1900, agriculture employed 55% of the civilian labor force. Now, about 2%, depending how you count seasonal labor. In 1970 manufacturing was about 35% (IIRC) of US employment. Now, 14% and falling. In 2011, services employ over 70% of the labor force. In 2041, ..?
In services the big occupation is "cashier", with "retail supervisor" next (again, IIRC. Check the Bureau of Labor Statistics website if interested.) A comment up-thread talked about bus- and taxi-drivers, but truck-driving employs many more people - it's in the top ten of occupations.
It's not too hard to imagine these three occupations being automated. The first supermarket chain that figures out how to do for food what Amazon has done for books is going to make a killing. (Speaking of Amazon, automating its warehouses can't be more than a decade off.) And people are buying more and more take-out food, clothes and accessories on-line.
What can displaced cashiers and truck drivers do for income? That's where I get stuck.
The assumptions made by the people that assert "there will be jobs" are these:
1. Robot vision and statistical machine learning are dead ends and won't develop any further.
2. The space of human desires is unbounded.
3. Price elasticity of demand in aggregate is greater than one.
4. We will reintroduce checks to the tendency of markets to concentrate wealth and income.
Assumptions 2, 3, and 4 are of unknown validity. Assumption 1 is almost certainly wrong.
Where?
Indeed.
Little-known fact: the US has around 250,000 producing wells. However, the vast majority produce less than 10 barrels per day. IIRC the total number of wells drilled is over 400,000.
With 2,000,000 wells each producing 10 barrels, the USA will be set - provided that electricity stays cheap enough to operate all those well pumps. At present there are 1200 drilling rigs which can each drill two wells per year. So assuming no dry holes and indefinite well life, in 729 years: energy security!
Static analyses are great fun, but they are misleading.
PV cells have been decreasing in cost at the rate of 22% per doubling of production capacity for three decades now. There are good reasons to think that this trend will continue. Since PV now provides less than 0.03% of global energy, there's plenty of room for ten more doublings. That gets us down to under ten percent of current PV costs.
Balance-of-system costs (inverters, support structures, installation costs, and especially permits/approvals) have decreased more slowly than PV costs in the past, primarily because they used to be negligible. Now they are about the same size as PV cost, and lo and behold, people are starting to work on getting them down.
Now, details.
"1 ton of coal costs $36 = $0.006 per KWH"
So why is it that coal-fired power stations don't charge any less than $0.04 per kWh? Highway robbery!
You are -- or rather Green Econ is, please put quote markers on your quotes, and quote them properly -- comparing the cost of the coal to the capital cost of a PV plant, not the cost of its fuel, sunlight. The correct comparison is the capital costs of coal mines, railroads, and power stations versus the capital costs of PV installations. Capital cost is why coal-fired power stations charge 0.04, and they couldn't charge much less if the coal was free.
Green Econ is a shill for the coal industry, but only fools the uncritical. Good critical thinking practice for you!
"Historical records have had the average temperature warmer than now"
That may be true in another world of the many-worlds multiverse. Not this one.
Not globally, not even for the Northern Hemisphere, not for any climatically meaningful interpretation of "now".
Really, people, this is not hard. Google for Spencer Weart, read his website, then google Skeptical Science and read John Cook's web site.
"the ice caps . . . just three years ago, people were claiming they'd be gone in six [years]..."
Aha. I see the problem: reading comprehension. It was not ice caps but Arctic sea ice that was exercising the imaginations of bloggers. Cryosphere researchers expect the ice caps to last thousands of years -- tens of thousands of years, in the case of the Antarctic ice cap.
Humans need energy too, you know. Yes, the chain of work from sun->plants->other stuff->humans->human labor is highly energy efficient, but do you really think it's impossible to approach that level of efficiency artificially?
Kidding, right? Sun->plants: about 0.5% in-the-field efficiency for the harvested bits, at best. Plants->other stuff: about 70% efficient for harvest losses, transport and storage losses, and pre-sale losses. Oh, you wanted milk, eggs or meat? 10% efficiency on average. Other stuff->humans: about 60% to 80% efficient. Humans->human labor: maximum eight-hour rate, about 10% efficient. Overall: 0.05 * 0.7 *0.5 * 0.8 *0.1 = 0.14%. One part in one-thousandth.
Yes, I think we might be able to do as well as that with photovoltaic panels and electric motors.
What Stoney says. Especially point 1.
If you're past point 1, then at the risk of starting a flame war:-
For a compute-bound problem that *can't go on GPUs* and needs fast turnaround, you'd be silly to use anything other than Sandy Bridge at the moment. Core i7 2600 if you can tolerate consumer grade (keep a spare) or the equivalent Xeon, or better. (See, I told you, flames! Look, guys, I like AMD, I want them to succeed, but benchmarks are benchmarks.) For the combination of raw speed, FLOPS/watt, and minimum idle power, Sandy Bridge beats everything else at present. If you're paying for the power, you'll maximise your total computation per pound with SNB. And be much more likely to have your overnight jobs finished and waiting for you when you come in in the morning.
Regarding second-hand stuff, even if you can get it for free, if it's older/smaller than Core 2 Quad Q6600 or the equivalent Xeon, pass on it. There aren't enough cpu instructions executed per second or per kilowatt-hour in anything older. Some of the older cases can be re-used, though.
Spend the money on a programmer to parallelize the algorithm on standard CPUs, and put it out on BOINC.
No, DON'T use BOINC. It will take six months to set up the project and publicise it, and a further year to learn how to deal with the volunteers so as to get work done reasonably reliably. Half or more of the work you send out will never come back. Newbs will bother you with questions about BOINC and the sixteen other projects they're 'donating time' to, not your project. (Virus scanners cause people grief.) RAC hunters will badger you incessantly about minor discrepancies in points awarded.
Running a project through BOINC and getting value out of it means committing a *lot* of time and effort just on the PR side of the project-- setting up and maintaining a website and the Boinc infrastructure, and doing regular news updates, answering emails, getting rid of spammers, and so on.
Unless, of course, you don't get any volunteers.
Unless your work is very interesting to the general public and you can use a big PR machine, trying to use BOINC is a good way to waste a year and achieve nothing.
Better hope your manager doesn't read Slashdot. She'll be asking herself, "what use have I got for someone who takes weeks to realise he needs to use his editor's 'show white space characters' command when dealing with a white space sensitive language?"
It's a bit harsh, calling someone "dumbass" for not knowing something that's little known.
Sports analogy: programming in C++ forces you to pay attention to your breathing. A marathoner might want this; a volleyball player, not so much.
http://news.slashdot.org/story/10/12/27/1734204/California-Rare-Earth-Mine-Reopens
Aye, LXDE.
I have a Core i7 2600 and a Velociraptor, and LXDE is noticeably faster than Xfce, let alone Gnome 2 or KDE. So much so that I always go back to it, despite its comparative lack of features, and I'm no hyperactive multitasker.
Speed has been the number 1 user interface problem ever since the days of TTYs. And it probably always will be. I'm surprised someone like Torvalds isn't already using LXDE.
you can buy a near infinite amount of services and be environmentally sustainable.
Only if the services are completely intangible, for example buying a Vuitton branded plastic grocery bag for $1M rather than an unbranded one for $0.05. Nearly all services consume energy and other resources in the act, and competition keeps their price at the marginal cost of supply. We can only increase the amount of energy consumed about 100 times before heat pollution starts to get serious, and prices can't rise arbitrarily.
I keep wondering too. Haven't heard a satisfactory answer in five years of asking.
Some of the more sensible things that have been suggested: arts and entertainment - everybody can make money by being famous (yeah, right); sports - a subcategory of entertainment; personal services: be a lifestyle coach or language tutor or personal shopper; ... or ... uh, that's about it, really.
A job that I hope returns in large numbers: union delegate. But I expect to wait at least twenty years.
Specialization is a pretty basic to macroeconomics.
Yes! That explains why the Germans make cars and sell them to the French, and in return the French make ... cars, and ... and sell them to the, the ... the Germans...
Wait, what?
I think you mean, specialization is fundamental to increaasing productivity at the firm and industry level. That is, in microeconomics.
The jobs we outsource are jobs we don't actually want because they don't pay well enough and are pretty unskilled.
Of course. That explains why since the recession bottomed out, nearly all the growth in employment in the USA has been in jobs that pay less than the median wage. ... Oh, wait.
A $20K car from 1980 is certainly higher quality than a $20k car from today, too.
Actually, no. A vanilla Honda Civic of today has vastly better suspension, brakes, steering, and headlamps; it has side-intrusion beams and airbags; it's much slower to rust or fade; and it's much more reliable than a $20,000 luxury sports car of 1981. Its acceleration isn't quite as good, but it's not as far back as you might think.
This is leaving aside frills like electric windows, air conditioning, sound damping (so that you can actually hear the stereo), and remote central locking, some of which the older car may have had, briefly. The Civic won't have leather seats, a chrome-plated ashtray or a walnut knob on the stick shift, but I know which I'd prefer to use daily.
Extrapolating from cars, we should have much better value from sound systems than we do. The fact that we don't have any improvement is because most people don't care.
IMHO, there's nothing that sounds quite as nice as wood for a speaker enclosure
Concrete is better -- it has very little in the way of resonances at the energies and frequencies experienced in speakers. Your speakers sound more like the design calcs say they should. But we're way into DIY territory here, lining the insides of our speakers with an inch of concrete all 'round.
You're right that the speakers are far and away the most important determinant of the sound. The rule of thumb is, half your budget on the speakers, and the rest for everything else.
The human ear can hear up to 22KHz.
The young, female human ear can hear up to 22 kHz. (Off topic: why can't people use SI prefixes properly, goddammit? "Kilo" is "k", not "K". Mega is M, not m.)
The male ear over 20 can't hear much more than 19 kHz, considerably lower if it was brought up in an urban environment. And it's all downhill from there. By 50, you're lucky if you can hear much over 9 kHz.
Shannon’s Law says that a digital signal is indistinguishable from an analog signal at double the frequency of the analog signal
No, it doesn't.
Shannon's law says you can recover a sine wave from a sample at double the wave's frequency. Music and speech are not sine waves.
Hint: in professional sound recording, sampling is done at 192 kHz because sound engineers can hear a difference between that and 96 kHz, let alone {shudder} 44.1 kHz. They don't do it for fun: the files are double the bulk, so the engineers have to truck around more storage and wait longer for files to transfer. Time is money. But they use 192 kHz nonetheless.