Extry! Extry! Read all about it! People Pay More for Higher Quality Merchandise!
Related Stories:
BMWs More Expensive Than Kias
Macs More Costly than PCs
Canon Cameras Sell For More Than Kodak Cameras
But seriously, why wouldn't they be more expensive? You get a much, much nicer end product. Why would you pay $10 for a hamburger at Outback when you can get one for a dollar at Mickey D's? They both feed you (poorly!), but one is much more pleasant to eat than the other. How about a music file? Are you happy with a 64kbps encoding of a tune, or do you prefer a lossless encoded version?
It's the same with an HD movie -- it's much more pleasant to look at HD than an NTSC quality movie.
My biggest point is "Getting the job done" instead of "Getting the job done right".
People don't seem to know that there is a difference anymore, it makes me sad to think that in another 10 years when even more unqualified professionals get on the scene, my salary will go down even further.
There's hope for you. 10 years can be enough time for management to realize that some people do "get it", and that these people are far more valuable than the clicky-coder monkeys they've been picking up for cheap.
Of course, I'm an optimist and presume that not every manager is pointy-haired (or at least not all the time.)
Microsoft hasn't been pushing NGEN really hard to us. I was kind of surprised, but the.NET gurus have told them "do lots of testing before you NGEN your assemblies in the field." I'm not sure if it's arrogance on their part, certain that the.NET runtimes will be fast enough for everyone who doesn't know any better; or if they're trying to be cautious about people running invalid pre-jitted assemblies (yes, I know the cache is SUPPOSED to be invalidated if conditions change.)
Perhaps it's because NGEN is likely to be run exactly once at installation time. However, if your conditions change and your cache is invalidated, they'll pay the JIT performance when they run it afterwards, causing your users to wonder why their app slowed down even though they "didn't change anything, honest!"
It's not even an issue of "logical" hierarchy. It's one of "repeatability." Once I have learned that "Tools/Options" is where I go to change menu settings in Outlook, I expect "Tools/Options" to be there the next time I need to change menu settings. With the whole "auto-hide" stuff, if the Options item goes unclicked for a month or two, it will hide itself.
In six months I need to change menu settings again, so I vaguely recall "Tools/Options." I click "Tools", but now there is no "Options" choice. My brain says "ouch, John, you screwed up." So I begin hunting for Options elsewhere in the menu tree. Well, three versions ago I think they used to be under "Edit". Nope, not there. "Actions"? Not there. Back to "Tools" again, and start looking at other choices like "Organize"? No, not organize. Oh, yeah, there's a down-arrow triangle thingy here, so I'll click that. Move mouse, and pause for exactly the wrong amount of time (410 ms) prior to clicking the down arrow. And in that 410 milliseconds, my hover action over the arrow caused all the previously hidden choices to spring forth! So just as I click, a random menu choice throws itself under my mouse pointer. F'ing brilliant, Microsoft, you've perfected the nadir of intelligent menu interface design.
So until they actually do make my life easier and simpler with "intelligence", perhaps they shouldn't be publishing books on the subject like they're some kind of experts.
Well, there's bloat and then there's bloat. And there's speed, and then there's speed.
A.NET app is "compiled" to MSIL, the intermediate token language common to.NET (comparable to java bytecode.) Tokens can be smaller than their machine language counterparts simply because they can be used to represent more complex ideas. And, as in java, the.NET libraries are quite extensive, encapsulating lots of functionality available to your application by reference. So MSIL should theoretically be smaller than a.EXE representing the same program, meaning it's not as bloated.
But then there's the 30MB distro of the runtime engine. And it has something like a 10MB footprint on your local machine's RAM, not counting your application. So there's your bloat.
As for speed, the runtime environment is smart. When you install it, it looks around your box and says "Hey, you have an Athlon64 4000+, I have all these optimized instructions here for doing for loops, while statements, etc." So each machine has its own unique runtime environment that is optimized for its CPU, drive resources, etc. The app developer doesn't have to ship "lowest common denominator 80386 code." And he doesn't have to ship "fat binaries" to send optimized code containing the differences between Pentiums, Xeons, or Athlons. The code your machine ultimately executes is going to be very close to optimal for your equipment.
But then again, running a just-in-time compiler means that every single time your application loads up, it's going to dog it until everything's been compiled. JIT is fine for long running apps, but sucks for the transient jobs. Plus, with a giant RAM footprint comes giant amounts of swapping. No matter what you do on a box, it's always slower when it has to swap in some RAM first.
So, is.NET "bloated" or "slow"? I think that decision hinges more on your application and intended usage.
Microsoft has long been wanting to push AI in their products. That's exactly what Microsoft Bob was supposed to be, remember?
I think a more likely reaction is: do users even want intelligence? As for me, I value "predictability" more than "intelligence." If I click the "Tools" menu in Outlook, I sure want to see "Options" listed below there, even if I haven't used it before. All this crap they've put in applications like Office to "hide" features I haven't used recently makes the menus far harder for me to use. The first thing I usually do is hunt around for the option menu to turn off the "auto-hide" function. The rest comes easy after that.
I think Microsoft needs to get things right before they make them smarter.
You're right, that is easy to say until you plunk down many moneys on a large plasma HDTV. The difference between SD and HD is almost one of magnitude, not degrees. Their idea, (which does not surprise me in the least) is that if you're willing to spend $xxxx on a TV you're both willing and able to spend an additional $10/month for a digital output cable box with DVR. (it's $5/month extra for the digital box with no DVR.) And so I am.
Yes, it blows, and it blows hard. But they're pretty much a monopoly -- you get a very similar craptacular interface from all the satellite decoder boxes I've seen, and I'd still have to pay the cable company for my internet feed anyway.
Anyway, I'm content to blame the troubles on DRM and the psychotic protectionism -- the cable company is merely capitalizing on someone else's greed.
The DIY crowd will just record off the analog out, it's really at the "good enough" state anyway.
Not an option. My cable box doesn't have an analog out for HDTV. It has an HDCP stream coming over the HDMI cable.
I was perfectly happy with my analog-based ReplayTV, but it was relegated to the "little" TV once I got an HDTV set. I had to pay to lease the Comcast DVR cable box since it's the only game in town for recording HDTV. However, it sucks -- the software is more buggy than the ReplayTV software ever was, and the interface is much less friendly than the ReplayTVs. Plus, with the cable company firmly in control, they don't let me do things like "hide" unwanted channels. Like I'm ever going to watch QVC, or why I'd want to skip over 60 pay-per-view sports slots that I'll never watch?
Yes, my TV has a CableCard slot, and yes, it's most likely going to sit there unused forever. I want a DVR more than I want "one less component".
I'm thinking of buying an HDCP decoder so I can build my own DVR using the cable box anyway. But those decoders are still about 400 euros.
Firefox frequently goes on a memory eating binge for me. I've not strongly blamed Firefox in the past because I usually have about a dozen extensions loaded up, and so I've never had direct evidence that it's Firefox's fault vs. one of the extensions.
I do think it's related to the pages I visit -- for example, I notice it more after browsing a dozen photoshop forums on Fark. And a simple shutdown/startup does clean it up, of course.
I had thought it was PDF related, so I installed the PDF Download extension which allows you to open PDFs in a separate executable. That's helped a lot, and is part of the reason I haven't been keeping an eagle-eye on my memory consumption as of late.
But yeah, this has been an ongoing thing for me since at least Firefox 1.0.
Yes, and we all know that the dawn of time occurred precisely at midnight (GMT) on January 1, 1970. While it's rumored that time existed prior to then, there is no evidence for it in the syslogs.
Sorry but the rules of NIMBY say that if you've got the fewest neighbors, you lose. (Actually, the rules of the U.S. government are somewhat less restrictive: if you've got the fewest affluent white neighbors, you lose.)
Anyway, why not fill a valley near a long-term storage facility (like Yucca Mountain) with reactors? Designate the place "Official U.S. Government Nuclear Wasteland" and lease out space to all the power generators who want to build there? I know it's not what you want to hear, but development happens. Ask anyone whose childhood farmhome was displaced or surrounded by a suburban housing development. (Or ask a homeowner whose house is suddenly under the flight path of a new runway, thank you very much M.A.C!) What we should consider is to concentrate these industries into a single area: a nuclear plant isn't likely to complain about the "nuke plant being built next door."
At least that way it will affect the fewest back yards.
Oh, I fully agree with you. The risks of the different energy sources require that we employ all kinds of generation methods (including nuclear.) And I know that producing hydrogen will take a tremendous amount of energy. I definitely consider the "storability" of hydrogen to be one of its most important assets. Wind, tidal and solar power generators are all obviously variable in the amount they can produce, so storage of excess capacity is really a smart way to use them. And nuclear plants that are tied to hydrogen production could use "off-peak" capacity for production.
Land is becoming more and more scarce
The Minnesota analysis figured current land usage into their wind capacity planning. Modern turbines actually require only about 2% of the land around them (access roads, the tower itself, and powerlines.) That leaves 98% of the land around them still farmable. And the farmers' typically see much higher financial returns on their leasing of land to turbine owners than they do on their own crops.
There's a lot to be said for wind power -- except when it's not blowing. That's where storage becomes important, and there currently are no efficient storage mechanisms for large amounts of electrical power. All of them require conversion of the electricity to a physical form (all the ones I'm aware of that are suitable for megawatt power generation involve chemical reactions.) So if storage is a desirable attribute, why not as hydrogen for direct use in vehicles?
Hydrogen in liquid form has 10% of the density of gasoline.
"Lower density" means "less mass per volume" to me, but it doesn't indicate potential energy per unit of volume, so this comparison doesn't measure apples to apples. Did you mean to say "10% energy density"? Or would you have figures on energy density that you could post?
I get the Daimler-Chrysler "High Tech Report" annually, and they've been tracking their development of fuel cell vehicles for the last 10 years. One of their original goals was to use a liquid form of energy, such as methanol, because the mechanisms already exist for delivering volatile liquids to consumers. They seem to have backed off this, since the most recent issue describes their new "gas pump" for their hydrogen-powered vehicle, and they mentioned their "smart connector" that lets car and pump talk to each other, so the pump knows exactly how much fuel to deliver. I think one of the reasons they're reverting to gaseous hydrogen is the conversion of methanol to hydrogen requires a considerable amount of platinum as a catalyst.
They're getting a 200 mile range in their tiny test vehicles, so apparently they're able to store enough energy in the form of gaseous hydrogen to compare to a normal gasoline tank. On a side note: Consumer Reports was just loaned a hydrogen car to test, but their test track was over 100 miles from the nearest hydrogen station:-) (They chose to test it on local roads rather than truck it to their track.)
Nuclear generation isn't the only answer. A recent study here in Minnesota showed that we have enough wind resources here to provide 14 times our current electric consumption. That is, if we built all the windmills that they say we could build. And yes, the greenies are drooling all over these figures.
Right now, we have approximately 800 megawatt-hours generated in this state by wind turbines. That's the equivalent of one or two coal-fired electric plants. Our problem right now is one of distribution -- we barely have enough capacity to carry this energy now from the windy part of the state to the Minneapolis/St. Paul region. There are plans underway to build more distribution lines, but those always take a long time and stir up controversy.
Of course, this doesn't take into consideration anything to do with current oil- or natural gas-based consumption. As you said, converting any significant percentage of vehicles to hydrogen fuels would obviously require massive amounts of power we're not yet generating. And it takes lots of time and money to physically erect more turbines. But there's a lot less regulation required and a lot less complaining about wind power vs. nuclear power.
One advantage to hydrogen is that it could be shipped via truck. Rather than invest in thousands of miles of costly transmission lines (and pay the 10% power cost in transmission losses), hydrogen plants could be built next to new reactors located out in the Nevada and Utah deserts, and the hydrogen trucked to market.
Correct, that's covered by basic property rights. I own my machine, it's property.
Yes. But that doesn't mean the manufacturer has to give you a way to install the changed code. He doesn't have to provide socketed ROM chips, he doesn't have to give you an RS-232 jack, he doesn't have to give you anything. If you want to change it, you are likewise free to break out the soldering iron and figure it out for yourself.
But take that one step further. The manufacturer has every right to try to make it tough for you to change the software. He can embed the chips in epoxy. He can put tripwires inside the chip cases that will erase the chips if the case is broken. He can even add another epoxied-shut crypto chip that verifies the signature of the ROM chip.
And NONE of that has anything to do with the GPL. And it shouldn't. You want a nice, changeable machine? Don't buy one with a crypto chip embedded in epoxy, and don't support the manufacturers who resort to these tricks.
If you protect with DRM and use DMCA it may very well be illegal to change to code.
No, that's not even permitted by the GPL v2 today. The GPL explicitly states that they have to provide the source code, plus any changes they've made to the GPL'd code. You further are explicitly granted the right to take that code they've given you and modify it yourself. It does not address anything about where you can then run that code.
Trusted Platform plus DRM are basically trying to be the software equivalent of burning the code to RAM. That makes it the virutal equivalent of a hardware issue, and has almost nothing to do with the software running in that hardware. GPL v3 is trying to say "no no, don't let my code run on a Trusted Platform machine." Well, what's going to happen when every motherboard sold contains a trusted platform chip? GPL v3 code will be quietly squeezed out of existence.
GPL v2 works fine. It doesn't need tampering, and it doesn't need to arm-wrestle with the Trusted Computing issue.
I am totally against the way CORPOPRATIONS are using DRM technology
Yet they're the only ones who want it or have need of it. There are no "legitimate" end-user applications of DRM that wouldn't be better served by access controls, encryption and/or tamper-proof hardware.
As a private owner of computing resources, you may have need for security. Fine. That's what encryption is for. Need to restrict access to a hard drive? Use access control mechanisms. Need to prevent spies from copying data? Epoxy the USB ports shut (that's what certain agencies do.)
But trying to control "some" usage of a document? There are really two choices: use application level controls, such as turning on a NO-PRINT flag so your co-workers can't print it and sell it to the competition (and accept the risk that someone might figure out a way around the flag.) Or really encrypt it, and give only trusted users access to the decryption keys (or to the one secure machine with the file on it.)
DRM is just a shell game. It makes you think that you've encrypted it, but it's little more than turning on the NO-PRINT flag.
If the code that you wrote under the GPL was digitally signed by a manufacturer and put in a "Trusted Platform" machine (let's say a router), the GPL (v2) clearly states that they must make your code available, along with any modifications to your code.
That's it. The GPL grants you the further rights to take that modified code and change it any way you like. But it does not grant you the right to install that modified software back on that same machine.
This is no different than if that manufacturer took your GPLed code and burned into a ROM or blew it onto an FPGA and then soldered it into the router. Just because you wrote it doesn't mean the manufacturer has to give you provisions to alter it on their machine.
The answer is purely economic -- don't buy a trusted platform based machine. Don't buy an OS that supports trusted platforms (Vista.) Don't allow friends, families or your business to buy trusted platform machines. If you're in a position to purchase hardware, get "no hardware enforcement of digital signatures" written as a requirement into your RFQs.
GPL v3 is not a business friendly license. It will restrict development unnecessarily, and drive manufacturers away from using GPLv3 code. Look at how well Linux has done in embedded devices (the slug as well as many other Linksys appliances.) These restrictions will likely drive those manufacturers to alternate sources of software, or limit them to only GPL v2 versions of the code.
Actually, this virus might BE the "patch" for stupidity.
"Hey, what happened to all my documents?"
"You opened a pr0n attachment in your email, you just got what you deserved."
"Boy, I'll never do that again!"
So, if these idiots are capable of any learning at all, this might work out to be a good learning experience for them. And if they're not, well, hey -- it's not my problem they're stupid.
No, it doesn't matter if they post it publicly or not. My point was these are NOT American citizens, they're standing in their own lands, watching big shiny American satellites floating over their heads and reporting them. The satellites aren't exactly secret, and they're not in U.S. territory. The U.S. classification of information is not applicable to them.
This is where the concept of "Security through obscurity" is proven wrong yet again. If the NRO believed that these satellites were somehow "secret" just because NORAD didn't publish TLEs for them, a couple of Canadians, Australians and Germans have proven them wrong. And if these amateurs can do it with nothing more than a GPS receiver, binoculars and some pretty simple trigonometry, you can certainly believe that India, Pakistan, China, Iran, and Al Qaeda are all capable of the same feat.
Remember, it was astronomers who developed algebra and trigonometry over a thousand years ago to solve orbital problems. The math is not secret. The birds are not secret. The observers are not subject to any laws. There is nothing to be concerned about.
Umm, these are only secret U.S. satellites. Molczan is Canadian. Heavens-above is a German site. The observers are scattered around the world.
And in case you've only been listening to George Bush's version of world events, here's a news flash: the rest of the world doesn't give a rat's ass about spilling U.S. secrets. Most of them find it hilarious that a couple of guys with GPS receivers and binoculars can crack some of the secrets of the most expensive devices ever launched by the most powerful government on earth.
The point of warrants is in part to make sure that people cannot go around nicking things by pretending to be police.
Nice try, but Troll Tuesday was yesterday.
The point of warrants is that the constitution recognized that the police have more power than ordinary citizens, and that power had to be kept in check by someone else. In the case of warrants that's the court, and more specifically an elected judge. If the cops become "too powerful" and start abusing their search privileges, we the people can elect a new judge who won't issue so many warrants.
At least that's the idea they teach in Civics 101. These days, warrants are these things that we don't need so much if there are terra'rists running about. Just wiretap 'em all and let God sort out the infidels.
Better. What if they let go'a'uld characters "infest" other characters? Spy on them for a while, then slip back into another body and gate out! The SGC player wouldn't even know he'd been part of a plot.
Because all the other technology players have had their own flavor of OS to pimp. Sun has Solaris, IBM has AIX, HP has HP/UX, etc. Google is one of the few major "technology" players that has no ties to any particular major technology!
So Google has a blank slate, and the market is eagerly waiting for them to write on it. That is why the distro is likely to gain a lot of initial followers.
Along with the Google name, however, comes epic amounts of bandwidth and storage. You'll quickly realize you'll need only one host for apt-get: googlix.google.com. Users will come to expect everything will be there, and that everything they make available will work together as configured.
And that's going to push the long term acceptance way up.
This distro is likely to be successful simply because Google says it will be. And yeah, that is a bit frightening.
Slashdot Mirror
Related Stories:
But seriously, why wouldn't they be more expensive? You get a much, much nicer end product. Why would you pay $10 for a hamburger at Outback when you can get one for a dollar at Mickey D's? They both feed you (poorly!), but one is much more pleasant to eat than the other. How about a music file? Are you happy with a 64kbps encoding of a tune, or do you prefer a lossless encoded version?
It's the same with an HD movie -- it's much more pleasant to look at HD than an NTSC quality movie.
People don't seem to know that there is a difference anymore, it makes me sad to think that in another 10 years when even more unqualified professionals get on the scene, my salary will go down even further.
There's hope for you. 10 years can be enough time for management to realize that some people do "get it", and that these people are far more valuable than the clicky-coder monkeys they've been picking up for cheap.
Of course, I'm an optimist and presume that not every manager is pointy-haired (or at least not all the time.)
Perhaps it's because NGEN is likely to be run exactly once at installation time. However, if your conditions change and your cache is invalidated, they'll pay the JIT performance when they run it afterwards, causing your users to wonder why their app slowed down even though they "didn't change anything, honest!"
In six months I need to change menu settings again, so I vaguely recall "Tools/Options." I click "Tools", but now there is no "Options" choice. My brain says "ouch, John, you screwed up." So I begin hunting for Options elsewhere in the menu tree. Well, three versions ago I think they used to be under "Edit". Nope, not there. "Actions"? Not there. Back to "Tools" again, and start looking at other choices like "Organize"? No, not organize. Oh, yeah, there's a down-arrow triangle thingy here, so I'll click that. Move mouse, and pause for exactly the wrong amount of time (410 ms) prior to clicking the down arrow. And in that 410 milliseconds, my hover action over the arrow caused all the previously hidden choices to spring forth! So just as I click, a random menu choice throws itself under my mouse pointer. F'ing brilliant, Microsoft, you've perfected the nadir of intelligent menu interface design.
So until they actually do make my life easier and simpler with "intelligence", perhaps they shouldn't be publishing books on the subject like they're some kind of experts.
A .NET app is "compiled" to MSIL, the intermediate token language common to .NET (comparable to java bytecode.) Tokens can be smaller than their machine language counterparts simply because they can be used to represent more complex ideas. And, as in java, the .NET libraries are quite extensive, encapsulating lots of functionality available to your application by reference. So MSIL should theoretically be smaller than a .EXE representing the same program, meaning it's not as bloated.
But then there's the 30MB distro of the runtime engine. And it has something like a 10MB footprint on your local machine's RAM, not counting your application. So there's your bloat.
As for speed, the runtime environment is smart. When you install it, it looks around your box and says "Hey, you have an Athlon64 4000+, I have all these optimized instructions here for doing for loops, while statements, etc." So each machine has its own unique runtime environment that is optimized for its CPU, drive resources, etc. The app developer doesn't have to ship "lowest common denominator 80386 code." And he doesn't have to ship "fat binaries" to send optimized code containing the differences between Pentiums, Xeons, or Athlons. The code your machine ultimately executes is going to be very close to optimal for your equipment.
But then again, running a just-in-time compiler means that every single time your application loads up, it's going to dog it until everything's been compiled. JIT is fine for long running apps, but sucks for the transient jobs. Plus, with a giant RAM footprint comes giant amounts of swapping. No matter what you do on a box, it's always slower when it has to swap in some RAM first.
So, is .NET "bloated" or "slow"? I think that decision hinges more on your application and intended usage.
I think a more likely reaction is: do users even want intelligence? As for me, I value "predictability" more than "intelligence." If I click the "Tools" menu in Outlook, I sure want to see "Options" listed below there, even if I haven't used it before. All this crap they've put in applications like Office to "hide" features I haven't used recently makes the menus far harder for me to use. The first thing I usually do is hunt around for the option menu to turn off the "auto-hide" function. The rest comes easy after that.
I think Microsoft needs to get things right before they make them smarter.
Sorry. I saw one on the web back in December, but I'll be damned if I can find the link now. I *knew* I should have bought it when I saw it!
Yes, it blows, and it blows hard. But they're pretty much a monopoly -- you get a very similar craptacular interface from all the satellite decoder boxes I've seen, and I'd still have to pay the cable company for my internet feed anyway.
Anyway, I'm content to blame the troubles on DRM and the psychotic protectionism -- the cable company is merely capitalizing on someone else's greed.
Not an option. My cable box doesn't have an analog out for HDTV. It has an HDCP stream coming over the HDMI cable.
I was perfectly happy with my analog-based ReplayTV, but it was relegated to the "little" TV once I got an HDTV set. I had to pay to lease the Comcast DVR cable box since it's the only game in town for recording HDTV. However, it sucks -- the software is more buggy than the ReplayTV software ever was, and the interface is much less friendly than the ReplayTVs. Plus, with the cable company firmly in control, they don't let me do things like "hide" unwanted channels. Like I'm ever going to watch QVC, or why I'd want to skip over 60 pay-per-view sports slots that I'll never watch?
Yes, my TV has a CableCard slot, and yes, it's most likely going to sit there unused forever. I want a DVR more than I want "one less component".
I'm thinking of buying an HDCP decoder so I can build my own DVR using the cable box anyway. But those decoders are still about 400 euros.
I do think it's related to the pages I visit -- for example, I notice it more after browsing a dozen photoshop forums on Fark. And a simple shutdown/startup does clean it up, of course.
I had thought it was PDF related, so I installed the PDF Download extension which allows you to open PDFs in a separate executable. That's helped a lot, and is part of the reason I haven't been keeping an eagle-eye on my memory consumption as of late.
But yeah, this has been an ongoing thing for me since at least Firefox 1.0.
Yes, and we all know that the dawn of time occurred precisely at midnight (GMT) on January 1, 1970. While it's rumored that time existed prior to then, there is no evidence for it in the syslogs.
Anyway, why not fill a valley near a long-term storage facility (like Yucca Mountain) with reactors? Designate the place "Official U.S. Government Nuclear Wasteland" and lease out space to all the power generators who want to build there? I know it's not what you want to hear, but development happens. Ask anyone whose childhood farmhome was displaced or surrounded by a suburban housing development. (Or ask a homeowner whose house is suddenly under the flight path of a new runway, thank you very much M.A.C!) What we should consider is to concentrate these industries into a single area: a nuclear plant isn't likely to complain about the "nuke plant being built next door."
At least that way it will affect the fewest back yards.
Land is becoming more and more scarce
The Minnesota analysis figured current land usage into their wind capacity planning. Modern turbines actually require only about 2% of the land around them (access roads, the tower itself, and powerlines.) That leaves 98% of the land around them still farmable. And the farmers' typically see much higher financial returns on their leasing of land to turbine owners than they do on their own crops.
There's a lot to be said for wind power -- except when it's not blowing. That's where storage becomes important, and there currently are no efficient storage mechanisms for large amounts of electrical power. All of them require conversion of the electricity to a physical form (all the ones I'm aware of that are suitable for megawatt power generation involve chemical reactions.) So if storage is a desirable attribute, why not as hydrogen for direct use in vehicles?
"Lower density" means "less mass per volume" to me, but it doesn't indicate potential energy per unit of volume, so this comparison doesn't measure apples to apples. Did you mean to say "10% energy density"? Or would you have figures on energy density that you could post?
I get the Daimler-Chrysler "High Tech Report" annually, and they've been tracking their development of fuel cell vehicles for the last 10 years. One of their original goals was to use a liquid form of energy, such as methanol, because the mechanisms already exist for delivering volatile liquids to consumers. They seem to have backed off this, since the most recent issue describes their new "gas pump" for their hydrogen-powered vehicle, and they mentioned their "smart connector" that lets car and pump talk to each other, so the pump knows exactly how much fuel to deliver. I think one of the reasons they're reverting to gaseous hydrogen is the conversion of methanol to hydrogen requires a considerable amount of platinum as a catalyst.
They're getting a 200 mile range in their tiny test vehicles, so apparently they're able to store enough energy in the form of gaseous hydrogen to compare to a normal gasoline tank. On a side note: Consumer Reports was just loaned a hydrogen car to test, but their test track was over 100 miles from the nearest hydrogen station :-) (They chose to test it on local roads rather than truck it to their track.)
Right now, we have approximately 800 megawatt-hours generated in this state by wind turbines. That's the equivalent of one or two coal-fired electric plants. Our problem right now is one of distribution -- we barely have enough capacity to carry this energy now from the windy part of the state to the Minneapolis/St. Paul region. There are plans underway to build more distribution lines, but those always take a long time and stir up controversy.
Of course, this doesn't take into consideration anything to do with current oil- or natural gas-based consumption. As you said, converting any significant percentage of vehicles to hydrogen fuels would obviously require massive amounts of power we're not yet generating. And it takes lots of time and money to physically erect more turbines. But there's a lot less regulation required and a lot less complaining about wind power vs. nuclear power.
One advantage to hydrogen is that it could be shipped via truck. Rather than invest in thousands of miles of costly transmission lines (and pay the 10% power cost in transmission losses), hydrogen plants could be built next to new reactors located out in the Nevada and Utah deserts, and the hydrogen trucked to market.
Yes. But that doesn't mean the manufacturer has to give you a way to install the changed code. He doesn't have to provide socketed ROM chips, he doesn't have to give you an RS-232 jack, he doesn't have to give you anything. If you want to change it, you are likewise free to break out the soldering iron and figure it out for yourself.
But take that one step further. The manufacturer has every right to try to make it tough for you to change the software. He can embed the chips in epoxy. He can put tripwires inside the chip cases that will erase the chips if the case is broken. He can even add another epoxied-shut crypto chip that verifies the signature of the ROM chip.
And NONE of that has anything to do with the GPL. And it shouldn't. You want a nice, changeable machine? Don't buy one with a crypto chip embedded in epoxy, and don't support the manufacturers who resort to these tricks.
No, that's not even permitted by the GPL v2 today. The GPL explicitly states that they have to provide the source code, plus any changes they've made to the GPL'd code. You further are explicitly granted the right to take that code they've given you and modify it yourself. It does not address anything about where you can then run that code.
Trusted Platform plus DRM are basically trying to be the software equivalent of burning the code to RAM. That makes it the virutal equivalent of a hardware issue, and has almost nothing to do with the software running in that hardware. GPL v3 is trying to say "no no, don't let my code run on a Trusted Platform machine." Well, what's going to happen when every motherboard sold contains a trusted platform chip? GPL v3 code will be quietly squeezed out of existence.
GPL v2 works fine. It doesn't need tampering, and it doesn't need to arm-wrestle with the Trusted Computing issue.
Yet they're the only ones who want it or have need of it. There are no "legitimate" end-user applications of DRM that wouldn't be better served by access controls, encryption and/or tamper-proof hardware.
As a private owner of computing resources, you may have need for security. Fine. That's what encryption is for. Need to restrict access to a hard drive? Use access control mechanisms. Need to prevent spies from copying data? Epoxy the USB ports shut (that's what certain agencies do.)
But trying to control "some" usage of a document? There are really two choices: use application level controls, such as turning on a NO-PRINT flag so your co-workers can't print it and sell it to the competition (and accept the risk that someone might figure out a way around the flag.) Or really encrypt it, and give only trusted users access to the decryption keys (or to the one secure machine with the file on it.)
DRM is just a shell game. It makes you think that you've encrypted it, but it's little more than turning on the NO-PRINT flag.
That's it. The GPL grants you the further rights to take that modified code and change it any way you like. But it does not grant you the right to install that modified software back on that same machine.
This is no different than if that manufacturer took your GPLed code and burned into a ROM or blew it onto an FPGA and then soldered it into the router. Just because you wrote it doesn't mean the manufacturer has to give you provisions to alter it on their machine.
The answer is purely economic -- don't buy a trusted platform based machine. Don't buy an OS that supports trusted platforms (Vista.) Don't allow friends, families or your business to buy trusted platform machines. If you're in a position to purchase hardware, get "no hardware enforcement of digital signatures" written as a requirement into your RFQs.
GPL v3 is not a business friendly license. It will restrict development unnecessarily, and drive manufacturers away from using GPLv3 code. Look at how well Linux has done in embedded devices (the slug as well as many other Linksys appliances.) These restrictions will likely drive those manufacturers to alternate sources of software, or limit them to only GPL v2 versions of the code.
"Hey, what happened to all my documents?"
"You opened a pr0n attachment in your email, you just got what you deserved."
"Boy, I'll never do that again!"
So, if these idiots are capable of any learning at all, this might work out to be a good learning experience for them. And if they're not, well, hey -- it's not my problem they're stupid.
This is where the concept of "Security through obscurity" is proven wrong yet again. If the NRO believed that these satellites were somehow "secret" just because NORAD didn't publish TLEs for them, a couple of Canadians, Australians and Germans have proven them wrong. And if these amateurs can do it with nothing more than a GPS receiver, binoculars and some pretty simple trigonometry, you can certainly believe that India, Pakistan, China, Iran, and Al Qaeda are all capable of the same feat.
Remember, it was astronomers who developed algebra and trigonometry over a thousand years ago to solve orbital problems. The math is not secret. The birds are not secret. The observers are not subject to any laws. There is nothing to be concerned about.
And in case you've only been listening to George Bush's version of world events, here's a news flash: the rest of the world doesn't give a rat's ass about spilling U.S. secrets. Most of them find it hilarious that a couple of guys with GPS receivers and binoculars can crack some of the secrets of the most expensive devices ever launched by the most powerful government on earth.
Nice try, but Troll Tuesday was yesterday.
The point of warrants is that the constitution recognized that the police have more power than ordinary citizens, and that power had to be kept in check by someone else. In the case of warrants that's the court, and more specifically an elected judge. If the cops become "too powerful" and start abusing their search privileges, we the people can elect a new judge who won't issue so many warrants.
At least that's the idea they teach in Civics 101. These days, warrants are these things that we don't need so much if there are terra'rists running about. Just wiretap 'em all and let God sort out the infidels.
Better. What if they let go'a'uld characters "infest" other characters? Spy on them for a while, then slip back into another body and gate out! The SGC player wouldn't even know he'd been part of a plot.
Because all the other technology players have had their own flavor of OS to pimp. Sun has Solaris, IBM has AIX, HP has HP/UX, etc. Google is one of the few major "technology" players that has no ties to any particular major technology!
So Google has a blank slate, and the market is eagerly waiting for them to write on it. That is why the distro is likely to gain a lot of initial followers.
Along with the Google name, however, comes epic amounts of bandwidth and storage. You'll quickly realize you'll need only one host for apt-get: googlix.google.com. Users will come to expect everything will be there, and that everything they make available will work together as configured.
And that's going to push the long term acceptance way up.
This distro is likely to be successful simply because Google says it will be. And yeah, that is a bit frightening.