Sure. But sometimes people do it out of habit rather than because it makes sense. It aint a large sin, but it is somewhat silly to write cat file.txt | grep foo.
It doesn't make sense if it can be done safely, no.
A blanket law that says the equivalent of: "A blind person may never fire a weapon in a hunt." is stupid.
Laws should forbid that which you actually want to outlaw. The law should be general. Offcourse it should be forbidden to fire a weapon without adequate surety that it won't hurt a person or property. But the *means* for obtaining that knowledge shouldn't be specified.
It doesn't matter. We care that people don't go firing guns without knowing who or what they migth hit. We *don't* care from where this knowledge comes.
Checking a firing-range thoroughly, securing the perimeter, and thereafther darkening the room and attempting to hit a target by aiming for the noise it produces -- for example -- can be done perfectly safely, and there's no reason to forbid it.
Firing a gun after someone else has carefully aimed it for you, can be done equally safely, and there's no reason to forbid it.
For what it's worth, SMART-enabled harddiscs provide some of this info (use "smartctl" under linux to access it). You do not get sector-by-sector read/write-count, nor max-writes-to-sector or max-reads-from-sector.
But you *do* get numbers of hours powered up, total writes, and total reads. Which tells you something.
The laptop I'm typing this on for example, comes with a 120GB harddisc. It's been powered up for 2713 hours (which corresponds to a full workyear plus running overnigth on some occasions because I ran some load-tests). In those 2713 hours, a total amount of 15904GB was written to disc, and a total of 807.201GB was read from disc.
So, this means the entire hard-disc has been written 100 times and read 6000 times.
Obviously, since there's no attempt at load-balancing use of the blocks, there'll be some blocks that are written a lot more than 100 times, and a lot of other blocks that are written a single time and which have never changed since then. On flash-discs there's internal circuitry that load-balances writes though, and in addition to that flash-filesystems do load-balancing on their own. It's not tricky to do. All you need to do is swap a heavily used block against a little used block, and maintain an internal table of the mappings once some blocks are much more used than others.
For laptops having no movable parts, small size and low power-consumption are aditional bonuses.
If the efficiency is too low, then it requires to much *area*. Land is not free. The size of your roof is not unlimited.
If the price/watt is too high, then it doesn't matter how efficient it is -- it's still never going to pay for itself. I actually think you're being overly optimistic in your estimates that current solar-panels pay for themselves in 5 years.
If they did, they'd be good investments. Life-span is around 20 years, but as you say they get somewhat less efficient with age. So, if they repaid themselves in 5 years, that'd probably mean they end up paying back triple cost in 20 years. That gives you a 6% interest-rate on your investment, which is decent (not spectacular, but decent) and the interest-rate is *better* if you assume energy-prices are going to rise. (A fair assumption I think)
Sadly, unless solar-cells have improved a lot I don't think payback in 5 years is realistic. Last time I did the math it was more like 10-12 years. (which means never in economic terms: you'd be better off sticking your money in the bank) You may be thinking of *energy-payback* (i.e. the time it takes for a solar-cell to produce the same amount of energy that was spent in manufacturing it) that used to be like 5 years, but is today actually more like 1-2 years.
True. Making tiny writes is bad for flash. But there's reasonably simple ways of avoiding that most of the time.
For example, assume you where using flash for *swap*, true one flash-block is significantly larger than a memory-page (typically 1K-4K) so if you swapped out a single page, you'd be wasting flash-life. So you don't *do* that then.
Modern computers have memory from several hundred megabytes upwards. If you're in a situation where memory is tigth and you need to swap out something to free up RAM, there is no reason whatsoever for not swapping out 128KB (or whatever the block-size of the flash-device is) all at once.
That'd give you 128KB of free memory, or 0.125% of the RAM in a typical new computer. Notice that you can still swap back *in* a single memory-page when/if it becomes demanded -- flash doesn't have a limit on read-cycles.
If you make tiny changes to a file, say changing a single byte, then you're rigth; flash-life would decrease rapidly. But I'm not aware of any workload that consists of doing a constant high-load of individual tiny writes to different files. What would that be ?
I didn't say Flash is perfect for *All* applications. I said it's not a practical concern for most workloads. I'm sure you can construct perverse usage-scenarios where it's not.
It depends. If you want to "go binary" you need to either remove those 15%, or get the people who contributed those 15% to agree to some other license.
Whenever you accept any code from other people, you need to pay attention to what rigths you get to that code. If you want to take the result proprietary -- you need to convince contributors they should send you patches under a license that allows you to do whatever the heck you want with them. I'm sure I'm not the only programmer who wouldn't even consider agreeing to that unless you paid me to do so.
Why should I (or anyone) spend free time improving your proprietary product ?
Sure. Hunting is, for most people in the western world a recreational activity. Even though it may also contribute something worthwhile to the household, you'd mostly be better of spending the same number of house working and then buying the meat.
That is completely beside the point.
Disabled should be allowed to participate in all activities where they can do so with acceptable risk. This doesn't sound all that risky to me. If it makes *sense* is a moot question -- we don't go around forbidding activities for the reason that they're useless. Being a free society is (among other things) about letting people decide for themselves what makes sense for them.
Sure, you can live happily all your days without every hunting, owning a dog or collecting stamps. But that is no reason for outlawing those activities.
But the "limited writes" of flash isn't a problem in the real world for anything outside of a few niche markets.
New flash-drives typically can sustain on the order of 1 million writes.
Balancing ensures that that pretty much means 1 million writes of the *entire* capacity.
So, a 1GB flash-drive will be broken by the time you've written 1GB*1million bytes to it. Which corresponds to constantly writing to it at a speed of 10MB/s 24 hours a day, 365 days a year for a little over 3 *years*.
If you only do this for 8 hours a day, then it's 10 *years*.
Very few drives are used that intensively anyway. Yes, there's a few in datacentres and so on. But for all normal laptops and desktops it's a non-issue. If your desktop is swapping constantly, 10MB/s, 365 days a year, 24 hours a day, you should get more RAM.
As capacity grows, the problem gets *smaller*.
If you had a 100GB flash, suitable for using as a hd-replacement in a laptop, then even if it could only sustain 100.000 writes, that would still be 30 years worth of round-the-clock 10MB/s writing. No laptop-drive is used that intensely.
Meanwhile, those of us that insisted on digital music being in some well-documented un-encumbered format such as mp3 or ogg (I know mp3 aren't completely free, but the difference is academic for most home-users) have a free choice between pretty much *any* portable music-player present or future.
It's not as if there's a lot of portable music-players on the market that fails to support mp3. Nor is it likely there *will* be in the next several years.
Yeah, if you where *really* paranoid you'd go with a FLAC-library, then you could transcode-and-transfer to whatever is the current popular format. In *practice* though you're pretty likely to be able to play that mp3-file 50 years from now.
Yeah. I agree. Pro photographers will still be needed. Perhaps less than before, but still needed.
For unpredictable events, it may be the amateurs "wins", because actually *having* a picture/film of something beats not having it, regardless of quality.
If someone had actually filmed the lorry-driver in the swiss-alps tunnel stopping and putting out a fire on his lorry *twice* before, extinguisher empty, and the lorry still burning, deciding to drive the thing out to avoid an inferno -- it wouldn't *matter* if the images where dark, smoky, shaking, out-of-focus, whatever. People love heroics and the spectacular.
Sorry. No go. I attacked you for arguing as if American situations is *all* that exists.
I have a European (Norwegian) POV. But I'd *never* automatically assume that everyone I run into on the net is automatically Norwegian, nor that what works here is universally valid. I'm from Norway, but I'm aware that the world is bigger than that. You don't seem to.
Even in the USA your claim is bullshit. It is perfectly possible for a family in USA to manage perfectly well with just a single car. No, not *all* families. It depends on the specifics, like where you live, where you work, what the kids do and so on. But you never asked about any of that. Instead you point-blank claimed that it can't be done.
It's nonsense in the USA, and it's even *more* nonsense for anyone with a wider perspective.
It's also interesting that you accused me of being too agressive, and now you're yelling around like a wronged 5-year-old.
We're planning on renewing our air-force by buying some new figther-planes, and it looks as if Eurofigther, SAS-Gripen and the JSF are the most likely candidates.
The first suggestion from the US was that we'd not even be allowed to *see* the sourcecode for the JSF under NDA. I think that may have gotten resolved, but being allowed to *change* anything is out of the question.
It's ridicolous. Why would any sovereign nation accept buying military material where they're *dependant* on a foreign power for even trivial bugfixes ?
First, that's only true if you count time from starting Linux, which typically is several seconds after turning the PC on. That a modern computer frequently uses 5-10 seconds before even *starting* the OS is not the fault of the OS though. But it is annoying.
Secondly, that's the time from starting Linux and until init starts. It's true -- the kernel is booted at this point, the rest is just starting of user-space processes. But that's not the way the user perceives it.
The user is more likely to care how long it takes from pressing the power-button and until he has a fully responsive desktop infront of him.
Nah. It's easy for the OS to figure out which pages are free, and which are just buffering stuff that's already on disk. All it needs to do is pretend a large memory-request came in just now, and swap was full. This means ignoring free memory, memory that is holding copies of pages *already* in swap, memory holding mmaped files and memory holding various buffers and caches.
It takes my 2GB linux-box on the order of 100ms to figure out that to hibernate, it only really needs to save 700MB to disc.
A *lot* quicker than writing 1300megabytes of unnneded crap to disk. (and then reading it back in later)
Memory is *extremely* much quicker than disc. Almost *any* amount of memory-trawling on the part of the OS is going to be quicker than going to disc. And it's not even much extra code -- the OS needs to be able to figure out how to free up memory anyway, it potentially needs to do that every time a malloc() comes in.
I *prefer* living in a country with relatively small differences between rich and poor. True it *does* mean that getting a haircut or a BigMac or a cinema-ticket is expensive, but it has many advantages, besides the altruistic ones.
Look at how the rich has to live in societies with a *large* difference, such as for example South-Africa.
Castles. Cameras. High walls. Gated communities. Life-guards for their kids. Armoured cars.
I consider it worth an extra buck for my BigMac to live somewhere where social tensions are more manageable.
Solar ain't the rigth solution *everywhere*. But it may be the rigth solution *somewhere*.
If you're in an area with lots of snow and lots of pine-needles falling on the roof, odds are your particular roof wouldn't be a good choice.
To answer your other questions:
Some solar-cells can take walking, but the cheapest ones can't. So you'd probably need to install them in "fields" small enough that they can be cleaned with a broom while standing "beside" them, which rules out covering *all* of your roof with them. This may change as technology improves. Today the most common choice is to mount them *on* the roof, with normal roofing underneath. There are solutions where they replace the normal roofing, but at the moment those still cost more than the normal-roofing+solar-cells solution.
Normal life-time for solar-cells is in the range of 20-30 years, which I think is in line with most other common types of roogfing.
Solar-cells can be cleaned with a water-hose and a broom without getting damaged. It tends to be easier to do it simply with one of them high-pressure water-thingies though.
There's 1 basic reason why this ain't happening on a large scale:
Presently the cost (including installation) of solar-cells-on-roof is *higher* than that of buying power from the grid. So it's a net money-loss to do it.
This can change if solar-cells gets cheaper (they certainly will) if power off the grid gets more expensive (also not unlikely), or if there's subsidies or tax-breaks from the government. (it's beens suggested solar-cells and installation should be VAT-free for example)
It may be digital shadow. But the shadow moves with the sun. So if you built a solar-plant such that it shadowed 2/3rds of the ground, it'd still shadow *different* 2/3rds of the ground at different times of the day/year. I'm honestly not sure that this would be all that bad for plants growing at ground-level.
A different problem is that unrestricted plant-growth would probably be bad for the solar-plant, which is a real problem. In a desert there ain't that much "unrestricted" plant-growth anyway though.
That all being said, I'm still convinced the damaging effects would be smallest by distributing the solar-cells in many small installations, preferably in areas already built-down.
What is the environmental downside of covering the parking-lots of America with a solar-cell-roof 3-4 meter up ?
What is the environmental downside of covering the roof on my house with solar-cells ?
I'm sure there *is* one. There's no such thing as a free lunch. I'm still pretty convinced the downside of solar can be minimized to the point where it's one of the more ecological friendly power-sources out there.
Sure. That's why I wrote: "assuming warm/sunny climate".
This is the case for much of USA for example, you guys use the most power on warm sunny days.
It's not the case for for example Alaska or Norway (where I live), so solar-panels won't have those advantages here. We're going to have to rely on something else. At the moment Norway is 95% supplied by hydro-power, which works out fine since we've got lots of mountains and lots of rain.
Different solutions are optimum for different geographical areas.
Yeah sure. Plant-derived oil is a renewable resource. Precisely like plant-derived alcohol.
I was talking about the kind of oil that is there, free for the taking, just pump it out of the ground. (I know it's not really free, but it's incredibly cheap -- current production-costs in Norway for example are about 10% of the selling-price of oil, it's pretty close to pumping money out of the ground)
As a Norwegian -- you can go to MIT without having a rich daddy, simply by being one of the top students.
You see, the government recognizes that there's some subjects where foreign universities may be better, and so give a stipendium to send a number of students to them. I know, because my friend (Hi Stig !) did.
I didn't display any "anti-American" anything. I reacted (strongly) negatively to the idea that one cannot reasonably have a family with kids without owning atleast 2 cars. It's your loss if you consider *that* argument universally American. Personally I'm convinced 90%+ of Americans do *not* agree with that.
btw, Slashdot may be in the USA -- but that doesn't mean, not by far, that only Americans read it, or only American POVs are valid.
73% is a large fraction. A monopoly is more damaging the higher the barriers to entry are, and the harder it is for the customers to select another provider.
If barriers to entry are high, and it's a lot of cost/hassle to change provider, then a monopolistic provider can demand a far to high price, or too onerous terms, and the customers still have no realistic choice but to accept it.
Microsoft Windows is a good example.
To enter the competition you need a complete modern operating-system and a large selection of software running on it. Which is a humongous investment. Linux just *may* have that by this point, but that was only possible trough cooperation of hundreds of organizations and thousands of individuals over literally decades.
But worse, selecting a different provider of OS is a large hassle and costs a lot. I don't mean that the OS needs to cost a lot, I mean that the process ends up costing a lot. Installation of OS and applications, training, conversion of documents, dealing with network-effects (Windows has the advantage that most of your customers, sub-contractors are likely to run it) etc.
It does look as if Linux has a fair chance of breaking trough despite this, but it's hard going.
Contrast with Google as a search-provider.
Changing your search-engine is literally 2 mouse-clicks if you're on Firefox. You migth need a few more to update your bookmarks etc. But it's still *tops* 10 minutes of work. You also don't require much retraining, there's details for the advanced users (do "site:.no" work in the new search-engine, or is the syntax different ?) but most are simple users who do nothing more than enter 1-3 common english words and pressing search -- which works without change in all search-engines.
Network-effects are also weak. It doesn't much *matter* if your customers still use Google and you switch to some other searchengine.
So yeah, Google is dominant in search. But no, I don't think they can abuse the position much without losing it.
If everyone had plenty of money, then the money is simply worth less.
It may be worth less, but it's certainly not worthless.
Money basically buys you one of 2 things: somebodys *time* or somebodys *product*.
If everyone in a country got twice as much money, a likely result is that buying somebodys *time* would cost double, so in this respect you'd be back where you started.
For example, your hairdresser would earn double like everybody else, so you migth be twice as rich, but a haircut costs double, so it's the same thing, really.
With products it's different -- because productivity makes a difference.
If a farmer produces 1000kgs of flour in a year, then a kg of flour will cost (atleast) 1/1000th of his yearly salary. If more modern techniques allow the same farmer to produce 50000kg of flour, then flour will cost only 1/50th.
This is why the tendency when countries get richer, assuming the distribution of wealth stays constant, is for services to get expensive relative to products.
Norway, for example, is a good example of a country where services cost a lot -- for the simple reason that those performing the services earn a lot.
Having cheap services is only possible if those performing the service earn little.
Slashdot Mirror
Sure. But sometimes people do it out of habit rather than because it makes sense. It aint a large sin, but it is somewhat silly to write cat file.txt | grep foo.
A blanket law that says the equivalent of: "A blind person may never fire a weapon in a hunt." is stupid.
Laws should forbid that which you actually want to outlaw. The law should be general. Offcourse it should be forbidden to fire a weapon without adequate surety that it won't hurt a person or property. But the *means* for obtaining that knowledge shouldn't be specified.
It doesn't matter. We care that people don't go firing guns without knowing who or what they migth hit. We *don't* care from where this knowledge comes.
Checking a firing-range thoroughly, securing the perimeter, and thereafther darkening the room and attempting to hit a target by aiming for the noise it produces -- for example -- can be done perfectly safely, and there's no reason to forbid it.
Firing a gun after someone else has carefully aimed it for you, can be done equally safely, and there's no reason to forbid it.
But you *do* get numbers of hours powered up, total writes, and total reads. Which tells you something.
The laptop I'm typing this on for example, comes with a 120GB harddisc. It's been powered up for 2713 hours (which corresponds to a full workyear plus running overnigth on some occasions because I ran some load-tests). In those 2713 hours, a total amount of 15904GB was written to disc, and a total of 807.201GB was read from disc.
So, this means the entire hard-disc has been written 100 times and read 6000 times.
Obviously, since there's no attempt at load-balancing use of the blocks, there'll be some blocks that are written a lot more than 100 times, and a lot of other blocks that are written a single time and which have never changed since then. On flash-discs there's internal circuitry that load-balances writes though, and in addition to that flash-filesystems do load-balancing on their own. It's not tricky to do. All you need to do is swap a heavily used block against a little used block, and maintain an internal table of the mappings once some blocks are much more used than others.
For laptops having no movable parts, small size and low power-consumption are aditional bonuses.
If the efficiency is too low, then it requires to much *area*. Land is not free. The size of your roof is not unlimited.
If the price/watt is too high, then it doesn't matter how efficient it is -- it's still never going to pay for itself. I actually think you're being overly optimistic in your estimates that current solar-panels pay for themselves in 5 years.
If they did, they'd be good investments. Life-span is around 20 years, but as you say they get somewhat less efficient with age. So, if they repaid themselves in 5 years, that'd probably mean they end up paying back triple cost in 20 years. That gives you a 6% interest-rate on your investment, which is decent (not spectacular, but decent) and the interest-rate is *better* if you assume energy-prices are going to rise. (A fair assumption I think)
Sadly, unless solar-cells have improved a lot I don't think payback in 5 years is realistic. Last time I did the math it was more like 10-12 years. (which means never in economic terms: you'd be better off sticking your money in the bank) You may be thinking of *energy-payback* (i.e. the time it takes for a solar-cell to produce the same amount of energy that was spent in manufacturing it) that used to be like 5 years, but is today actually more like 1-2 years.
For example, assume you where using flash for *swap*, true one flash-block is significantly larger than a memory-page (typically 1K-4K) so if you swapped out a single page, you'd be wasting flash-life. So you don't *do* that then.
Modern computers have memory from several hundred megabytes upwards. If you're in a situation where memory is tigth and you need to swap out something to free up RAM, there is no reason whatsoever for not swapping out 128KB (or whatever the block-size of the flash-device is) all at once.
That'd give you 128KB of free memory, or 0.125% of the RAM in a typical new computer. Notice that you can still swap back *in* a single memory-page when/if it becomes demanded -- flash doesn't have a limit on read-cycles.
If you make tiny changes to a file, say changing a single byte, then you're rigth; flash-life would decrease rapidly. But I'm not aware of any workload that consists of doing a constant high-load of individual tiny writes to different files. What would that be ?
I didn't say Flash is perfect for *All* applications. I said it's not a practical concern for most workloads. I'm sure you can construct perverse usage-scenarios where it's not.
Whenever you accept any code from other people, you need to pay attention to what rigths you get to that code. If you want to take the result proprietary -- you need to convince contributors they should send you patches under a license that allows you to do whatever the heck you want with them. I'm sure I'm not the only programmer who wouldn't even consider agreeing to that unless you paid me to do so.
Why should I (or anyone) spend free time improving your proprietary product ?
That is completely beside the point.
Disabled should be allowed to participate in all activities where they can do so with acceptable risk. This doesn't sound all that risky to me. If it makes *sense* is a moot question -- we don't go around forbidding activities for the reason that they're useless. Being a free society is (among other things) about letting people decide for themselves what makes sense for them.
Sure, you can live happily all your days without every hunting, owning a dog or collecting stamps. But that is no reason for outlawing those activities.
New flash-drives typically can sustain on the order of 1 million writes.
Balancing ensures that that pretty much means 1 million writes of the *entire* capacity.
So, a 1GB flash-drive will be broken by the time you've written 1GB*1million bytes to it. Which corresponds to constantly writing to it at a speed of 10MB/s 24 hours a day, 365 days a year for a little over 3 *years*.
If you only do this for 8 hours a day, then it's 10 *years*.
Very few drives are used that intensively anyway. Yes, there's a few in datacentres and so on. But for all normal laptops and desktops it's a non-issue. If your desktop is swapping constantly, 10MB/s, 365 days a year, 24 hours a day, you should get more RAM.
As capacity grows, the problem gets *smaller*.
If you had a 100GB flash, suitable for using as a hd-replacement in a laptop, then even if it could only sustain 100.000 writes, that would still be 30 years worth of round-the-clock 10MB/s writing. No laptop-drive is used that intensely.
It's not as if there's a lot of portable music-players on the market that fails to support mp3. Nor is it likely there *will* be in the next several years.
Yeah, if you where *really* paranoid you'd go with a FLAC-library, then you could transcode-and-transfer to whatever is the current popular format. In *practice* though you're pretty likely to be able to play that mp3-file 50 years from now.
For unpredictable events, it may be the amateurs "wins", because actually *having* a picture/film of something beats not having it, regardless of quality.
If someone had actually filmed the lorry-driver in the swiss-alps tunnel stopping and putting out a fire on his lorry *twice* before, extinguisher empty, and the lorry still burning, deciding to drive the thing out to avoid an inferno -- it wouldn't *matter* if the images where dark, smoky, shaking, out-of-focus, whatever. People love heroics and the spectacular.
I have a European (Norwegian) POV. But I'd *never* automatically assume that everyone I run into on the net is automatically Norwegian, nor that what works here is universally valid. I'm from Norway, but I'm aware that the world is bigger than that. You don't seem to.
Even in the USA your claim is bullshit. It is perfectly possible for a family in USA to manage perfectly well with just a single car. No, not *all* families. It depends on the specifics, like where you live, where you work, what the kids do and so on. But you never asked about any of that. Instead you point-blank claimed that it can't be done.
It's nonsense in the USA, and it's even *more* nonsense for anyone with a wider perspective.
It's also interesting that you accused me of being too agressive, and now you're yelling around like a wronged 5-year-old.
We're planning on renewing our air-force by buying some new figther-planes, and it looks as if Eurofigther, SAS-Gripen and the JSF are the most likely candidates.
The first suggestion from the US was that we'd not even be allowed to *see* the sourcecode for the JSF under NDA. I think that may have gotten resolved, but being allowed to *change* anything is out of the question.
It's ridicolous. Why would any sovereign nation accept buying military material where they're *dependant* on a foreign power for even trivial bugfixes ?
First, that's only true if you count time from starting Linux, which typically is several seconds after turning the PC on. That a modern computer frequently uses 5-10 seconds before even *starting* the OS is not the fault of the OS though. But it is annoying.
Secondly, that's the time from starting Linux and until init starts. It's true -- the kernel is booted at this point, the rest is just starting of user-space processes. But that's not the way the user perceives it.
The user is more likely to care how long it takes from pressing the power-button and until he has a fully responsive desktop infront of him.
It takes my 2GB linux-box on the order of 100ms to figure out that to hibernate, it only really needs to save 700MB to disc. A *lot* quicker than writing 1300megabytes of unnneded crap to disk. (and then reading it back in later)
Memory is *extremely* much quicker than disc. Almost *any* amount of memory-trawling on the part of the OS is going to be quicker than going to disc. And it's not even much extra code -- the OS needs to be able to figure out how to free up memory anyway, it potentially needs to do that every time a malloc() comes in.
I *prefer* living in a country with relatively small differences between rich and poor. True it *does* mean that getting a haircut or a BigMac or a cinema-ticket is expensive, but it has many advantages, besides the altruistic ones.
Look at how the rich has to live in societies with a *large* difference, such as for example South-Africa.
Castles. Cameras. High walls. Gated communities. Life-guards for their kids. Armoured cars.
I consider it worth an extra buck for my BigMac to live somewhere where social tensions are more manageable.
If you're in an area with lots of snow and lots of pine-needles falling on the roof, odds are your particular roof wouldn't be a good choice.
To answer your other questions:
Some solar-cells can take walking, but the cheapest ones can't. So you'd probably need to install them in "fields" small enough that they can be cleaned with a broom while standing "beside" them, which rules out covering *all* of your roof with them. This may change as technology improves. Today the most common choice is to mount them *on* the roof, with normal roofing underneath. There are solutions where they replace the normal roofing, but at the moment those still cost more than the normal-roofing+solar-cells solution.
Normal life-time for solar-cells is in the range of 20-30 years, which I think is in line with most other common types of roogfing.
Solar-cells can be cleaned with a water-hose and a broom without getting damaged. It tends to be easier to do it simply with one of them high-pressure water-thingies though.
There's 1 basic reason why this ain't happening on a large scale:
Presently the cost (including installation) of solar-cells-on-roof is *higher* than that of buying power from the grid. So it's a net money-loss to do it.
This can change if solar-cells gets cheaper (they certainly will) if power off the grid gets more expensive (also not unlikely), or if there's subsidies or tax-breaks from the government. (it's beens suggested solar-cells and installation should be VAT-free for example)
Ok. Good point. That could be fixed by installing a mirror-roof anyway. So cancel *that* particular "advantage" then. :-)
Telegraph road -- which I thougth was an unlikely candidate for a sax-interpretation, but there you go.
A different problem is that unrestricted plant-growth would probably be bad for the solar-plant, which is a real problem. In a desert there ain't that much "unrestricted" plant-growth anyway though.
That all being said, I'm still convinced the damaging effects would be smallest by distributing the solar-cells in many small installations, preferably in areas already built-down.
What is the environmental downside of covering the parking-lots of America with a solar-cell-roof 3-4 meter up ?
What is the environmental downside of covering the roof on my house with solar-cells ?
I'm sure there *is* one. There's no such thing as a free lunch. I'm still pretty convinced the downside of solar can be minimized to the point where it's one of the more ecological friendly power-sources out there.
This is the case for much of USA for example, you guys use the most power on warm sunny days.
It's not the case for for example Alaska or Norway (where I live), so solar-panels won't have those advantages here. We're going to have to rely on something else. At the moment Norway is 95% supplied by hydro-power, which works out fine since we've got lots of mountains and lots of rain.
Different solutions are optimum for different geographical areas.
I was talking about the kind of oil that is there, free for the taking, just pump it out of the ground. (I know it's not really free, but it's incredibly cheap -- current production-costs in Norway for example are about 10% of the selling-price of oil, it's pretty close to pumping money out of the ground)
You see, the government recognizes that there's some subjects where foreign universities may be better, and so give a stipendium to send a number of students to them. I know, because my friend (Hi Stig !) did.
I didn't display any "anti-American" anything. I reacted (strongly) negatively to the idea that one cannot reasonably have a family with kids without owning atleast 2 cars. It's your loss if you consider *that* argument universally American. Personally I'm convinced 90%+ of Americans do *not* agree with that.
btw, Slashdot may be in the USA -- but that doesn't mean, not by far, that only Americans read it, or only American POVs are valid.
I have however watched the show from DVD, and found it enjoyable. Not -oh-my-god- enjoyable, but better than most tv-series.
If barriers to entry are high, and it's a lot of cost/hassle to change provider, then a monopolistic provider can demand a far to high price, or too onerous terms, and the customers still have no realistic choice but to accept it.
Microsoft Windows is a good example.
To enter the competition you need a complete modern operating-system and a large selection of software running on it. Which is a humongous investment. Linux just *may* have that by this point, but that was only possible trough cooperation of hundreds of organizations and thousands of individuals over literally decades.
But worse, selecting a different provider of OS is a large hassle and costs a lot. I don't mean that the OS needs to cost a lot, I mean that the process ends up costing a lot. Installation of OS and applications, training, conversion of documents, dealing with network-effects (Windows has the advantage that most of your customers, sub-contractors are likely to run it) etc.
It does look as if Linux has a fair chance of breaking trough despite this, but it's hard going.
Contrast with Google as a search-provider.
Changing your search-engine is literally 2 mouse-clicks if you're on Firefox. You migth need a few more to update your bookmarks etc. But it's still *tops* 10 minutes of work. You also don't require much retraining, there's details for the advanced users (do "site:.no" work in the new search-engine, or is the syntax different ?) but most are simple users who do nothing more than enter 1-3 common english words and pressing search -- which works without change in all search-engines.
Network-effects are also weak. It doesn't much *matter* if your customers still use Google and you switch to some other searchengine.
So yeah, Google is dominant in search. But no, I don't think they can abuse the position much without losing it.
It may be worth less, but it's certainly not worthless.
Money basically buys you one of 2 things: somebodys *time* or somebodys *product*.
If everyone in a country got twice as much money, a likely result is that buying somebodys *time* would cost double, so in this respect you'd be back where you started.
For example, your hairdresser would earn double like everybody else, so you migth be twice as rich, but a haircut costs double, so it's the same thing, really.
With products it's different -- because productivity makes a difference.
If a farmer produces 1000kgs of flour in a year, then a kg of flour will cost (atleast) 1/1000th of his yearly salary. If more modern techniques allow the same farmer to produce 50000kg of flour, then flour will cost only 1/50th.
This is why the tendency when countries get richer, assuming the distribution of wealth stays constant, is for services to get expensive relative to products.
Norway, for example, is a good example of a country where services cost a lot -- for the simple reason that those performing the services earn a lot.
Having cheap services is only possible if those performing the service earn little.