Uh, if you wanted to go with the flow, why aren't you using Windows?
It is plodding along just fine as well, with it's only-slightly-restrictive licensing practices.
Licenses matter. Linux will always do fine on the GPLv2 - companies like Tivo will ensure it! Now, whether the rest of the free software world ends up leaving it behind is a different matter. It obviously isn't going anywhere anytime soon, but stranger things have happened.
Re:Science is prediction, not explaination
on
The Trouble with Physics
·
· Score: 3, Interesting
Well, to the average layman most of Einstein's theoretical work seems meaningless. But, when you build a photocopier anybody can see that there is something to it.
Ditto for everything else that physics has discovered. The value of the discoveries is appreciated when it is seen how these discoveries apply to the real world.
The issue with string theory is that while it is self-consistent, it seems like nobody is able to actually do anything useful with it, and to me that makes it an inadequate theory, because the proof is in the ability to apply the theory.
I can plot my movements for the entire day and fit them to a 47-degree polynomial with a decent level of error, and then wax philisophical about the general theory of human locomotion. And that would last about as long as it takes somebody to realize that five minutes after I publish the theory fails to account for my subsequent activities.
Given a complex enough equation you can fit any set of data. And given enough time you can even make that equation look "beautiful". What I want to know is how well it holds up six months from now without constant tweaking...
Actually, most foreign wars work like that. The British lost very few engagements in the US Revolutionary war. However, the war functioned to create lots of dead bodies in red uniforms, and that was not popular back home. The democratic people of Britain were scratching their heads over why exactly they were sending their soldiers overseas to kill a bunch of Americans who really just seemed to want to just be left alone.
I tend to agree that on a tactical and strategic level the war in Vietnam was successfully fought (not wisely fought, mind you, but even if we did manage to kill thousands of our own soldiers with dumb policies we still managed to get the job done on the battlefield in spite of ourselves). Now, the whole notion of limited war was dumb, and prevented the US from just cleaning up. Korea is a better example of what could have been achieved, but I'd hardly consider North Korea a great success story. And that brings up the bigger issue - if you want to get involved in foreign civil wars you're going to find that social change is a lot messier than just winning battles.
So, yes, in Vietnam the US probably didn't lose a single engagement, unless you count the decision to send troops in at all...
But if you captured it using a framegrabber it would be: 1920*1080*30fps*12bpp*255min*60s/min/8bytes/bit = 1.4E12 bytes - and that isn't including the audio.
Even compressed HD uses quite a bit of space, but my point was that capturing DVI wasn't practical at all, even if you could afford the hardware (unless you buy realtime compression hardware).
Yeah, if you have a few TB of disk space. That card doesn't compress the signal - so you're drinking from a firehose. A 1 hour HD program would consume about 300GB of disk space if you used that card. Sure, you could then compress it down to a few GB, but not in realtime (figure 2 hours processing time for each 1 hour of recording time).
Want dual-tuners? Better have a bank of 6+ drives! Not only do you have to buffer all that video until it gets compressed, but you also have to stream it all at an insane rate of about 1.5Gbps - your SouthBridge won't even handle that, so you'd need a fancy motherboard design or several computers.
Consumer HD recording from DVI will only be practical when hardware mpeg2/4 compression becomes affordable. Right now the equipment for that is rack-mounted and costs a very pretty penny (computers, DSPs, signal-processors, probably lots of fiber, etc). TV studios use it, and they have trouble getting it all right...
Just to add a little to the other reply - your mythtv stream is only a few megabits/sec. Certainly not 700Mb/s! You might be able to record at that rate with the right drive setup, but you'd fill your hard drive mighty fast (80 megabytes per second eats through space pretty quick).
Go ahead and check the mythtv-users list - this is a common topic. The hardware capable of compressing live HD is very expensive - studio gear. We're not talking Apollo-mission cutting edge, but even the TV studios have difficulty with live HD streams (I know somebody who works in the industry).
Anybody recording HD using myth is recording compressed MPEG2 - not DVI. In fact, there is a company that will mod your cable box with DVI/HDMI-only output to add a firewire port so that you can record the HD stream at a decent rate.
Well, if person on the top-500 list has two kids, and they get every dime of the cash (no taxes, etc), neither would probably end up on the top-500 list.
Problems of economic division don't really hinge on the top-500 list - we're only talking 500 people here. And I do think that inheritance does reduce the incentive to achieve, and reduces the effective spending of cash.
My feeling is that inheritances below $1 million per recipient should be moderately taxed, and anything above that should be taxed at around the 90% level. If you can't get started in life with $1 mil cash then you'll be bound for welfare no matter how much you're given. Go ahead and adjust the exemption for COL and all that. Even with a 90% tax Bill Gates's kids would still be billionares.
If somebody earns their pay they shouldn't be punished for their success. I'm OK with progressive taxation, but let's not get to crazy with that. Unearned income is a different story (although investments shouldn't be too heavily taxed or nobody would invest).
Yet, in a perfect system, the benefit would be proportional to productivity -- which isn't close to reality.
Pick a person at random. Determine how much THEY increased their productivity, and look at how much their wages have increased. It will probably be decently correllated.
Now, if a person's employer bought a fancy machine that made their employees more productive, then the employees have NOT increased their own productivity, and consequently they will see very little of the productivity increase in their income. People raise their OWN productivity by obtaining an education, inventing better processes, etc. You don't raise your OWN productivity by showing up to work and doing what you're told, and people who do this will never make the kind of money those who do otherwise earn.
Those who contribute the most to an enterprise are going to reap the rewards. Sure, there is some random variation - some people inherit money (I support estate taxes for this reason), and some people win the lottery (a good way to tax the poor without making them upset). For the most part, though, if you want to get ahead in society you have to do more than show up.
Uh, why would I or somebody else want to take a job that pays $8/hour? You could probably do better than that running an open-source project and soliciting donations, let alone something that is actually useful to society.
If you're making $8/hour, I'd recommend getting an education and finding something to do with your life. You can do better than that trading on ebay. And a minimal education doesn't have to cost money - read wikipedia or something and you'll figure out something useful to do with your life.
One indicator that's often used to assess poverty is the percentage of income that goes into housing: if you're paying 50% of your income just to keep your mortgage steady, you're worse off than someone who pays 40% of their income to a landlord instead of to the bank.
Uh, sure, but be careful about how you measure it.
Money paid to mortgage-INTEREST is a housing expense. Money paid to mortgage-PRINCIPAL is not - you're paying it to yourself.
If you're just talking size of payments, I'd much rather pay 50% of my income to a mortgage than 40% to rent for 30 years! When I'm done in 30 years I have a nice house to my name, and I can sell it for a good chunk of cash. If I rent I own nothing, and I'm still paying 40% of my income to continue to rent.
Buying isn't always better than renting, but most of the time it is. How do you think the landlords make their money?
As far as air conditioning / etc being cheap goes - it isn't so much the cost to buy the unit as the cost to operate it. If you can afford those kinds of electric bills then you aren't actually in horribly bad shape.
No matter how you slice it, poverty isn't a bucket of roses.
And good thing that! Otherwise everybody would just sit around being poor!
Relax - no matter what the plans are you'll be able to buy one for 1/5th of cost. Just search ebay a month after they come out. Kids in developed countries will "lose" them and turn a nice profit (since they were free to them to begin with).
Why they're looking to promote the gray market by not selling retail is beyond me...
Agreed. For all the cost of these 10,000 year storage facilities and all the argument that goes on, they could just station 5000 marines at a number of reactors and just sell the fuel to operators at subsidized rates (or charge a pretty penny for disposal of non-reclaimed waste). If you only reburn the waste at a half-dozen facilities and run them as highly secure installations then it is very unlikely terrorists would get in. Honestly, rent-a-cops are enough to deter most terrorists as long as you have locked doors and such to slow down entry/escape from a secured facility. But hey, we're talking nukes, so go ahead and spend a little more to put black copters and tanks outside - it has to be cheaper than Yucca Mountain.
Well, if the only concern was DoS or other non-customer-approved use there is a simple solution - just give customers a website allowing them to specify domains/netblocks that are exempt from QoS filters. If they can base their filters on who is paying them, they can certainly base it on what the customer wants.
Implement QoS in a way that any given customer on average gets the same bandwidth, but QoS bits can be used to prioritize traffic for a single customer, or to prioritize across multiple customers for short bursts only. Also allow end-users to buy higher levels of service (which is already something ISPs sell).
Make the end-user the customer - not the remote site.
You can avoid congestion the same way you've always done it: don't do too many things at once over the same connection.
Well, only if the congestion is at your DSL modem/etc. If the congestion is at the OC3 line leading out of your ISP you'll have issues no matter what you do. Right now ISPs avoid such congestion, since customers would complain. Once they start charging for QoS they'll be sure to stop expanding their bandwidth - otherwise they don't have anything to sell (who needs QoS when you aren't bandwidth-starved?). Then pay-to-play websites will get the same service they get now, and anybody else will get degraded service. If you are in an area like mine you'll only have one or two choices for an ISP, and both will probably be doing the same thing.
If the telecoms really just cared about congestion at your DSL modem they'd just stop filtering out the QoS bits and instead ignore them everywhere but on the two ends of your DSL line. You'd get all the benefits of QoS without any disruption to your neighbors - if this remains the main choke-point. No need to charge money for it.
Then why don't you do your neighbors a favor and try to start your own cable company. You'll see how well the friendly monopoly treats you...:)
If it were purely tradition there would be SOME town SOMEWHERE that has a couple of utility providers. The fact that this is unheard of indicates that this is likely a natural monopoly.
About the only areas you see competition with local utilities is stuff like WiFi and satellite-cable. And this is precisely because they get rid of the huge localized investments that make it impossible to compete with utilities. If such a technology existed for broadband I'd be less in favor of net neutrality. I've seen Verizon's idea of premium internet service - I once tried it out on my cell phone. 95% of the WAP sites out there weren't reachable - because they didn't pay to play.
I'd be happy if linux worked right with my mouse when I use my cheap KVM. It works fine with windows, but not with linux. Sure, I understand the issues with mice being stateless and all that, but obviously MS has a fix. On my desk I have one keyboard, one monitor, and two mice as a result. And no, I don't want to go out and spend $300 on a nicer unit...
1) I assume that the kernel maintainers won't accept maintenance work for this driver. It would be silly for them to volunteer to maintain everyone else's proprietary software.:-)
But it wouldn't be proprietary if you gave it to them...
2) Buying another OS. Is that really what you advocate? It kinda takes away Linux's reputation as a hobbyists tool if you are suggesting that people not use it if they need a custom driver.
Well, there isn't much point if they're going to keep it to themselves. How many hobbyists need to hang onto the code, and wouldn't donate it back to the kernel team?
And commercial closed-source drivers will be easier to maintain. It seems to me this option benefits everyone - even people you hate. Is this an example of cutting off one's nose to spite their face?
But only the closed-source vendors would be able to maintain it. Suppose I have a TNT2 card - do you think that Nvidia is going to release a driver for it when linux 2.8 comes out?
The idea is to get rid of the closed-source easy-out - which should result in more code being opened. Sure, some code just won't get used at all, but that puts vendors who refuse to open their code at a disadvantage, at least among linux users. And why should Nvidia profit by selling cards to linux users when they don't release their code? The drivers wouldn't work without the kernel, and the kernel was made GPL to force redistributers to release their sources. Now, the kernel won't go GPLv3 anytime soon, but other software will, and future GPL versions will only tighten some of the closed-source loopholes that currently exist.
It has worked pretty well so far. And if somebody wants to make their own middle-layer they can, just don't expect the kernel team to help you out.
First - I'm very sorry to hear about this tragedy - I honestly have difficult replying just because I can't really equate the seriousness of what befell your mother with the content of this post.
That said, my feeling is that the only reason the police even investigated this theft is due to the associated near-murder of your mother. As many others have indicated, if a crime is "just a theft" it often gets ignored.
The tragedy is that most likely the thug that nearly killed your mother probably stole something before then, and maybe in one of those cases they could have been tracked down (maybe they made a mistake). If they were punished for the minor crime of theft then maybe things would have turned out better for your mother.
In my opinion crimes like theft and vandalism should be treated very seriously - they are personal in nature and if somebody is willing to smash your windows they will eventually be willing to smash you. Even if only $100 is stolen the police should be willing to spend a good sum to apprehend the criminal - eventually it will pay off.
Or what if the guy was attacked by a mountain lion?
Or what if the guy hit a dog with a branch to fend it off and the owner went nuts and attacked him without a gun? He may have gotten 10 years but at least he is alive..
It is all about escalation and quality of life. When I was in school somebody spray-painted the outside wall of the school. The next morning the principal delivered a long sermon, and within about 24 hours the pain was sandblasted off - probably at a cost of $10k or more. However, the move was effective - no more sparypainting for as long as I was there. If the paint were left up I'm sure somebody would have added to it within a week. Sure, the paint doesn't cost anything, but it makes th school look like a ghetto and probably would tend to result in the students acting more like they lived in one.
If somebody steals a $250 stereo the police should spend $40k tracking the guy down. Then fine and/or imprison the criminal for the full cost of the investigation - if they're underage put a lean on the parent's house. A town with a theft problem probably has only 10 thieves in it, and if you got rid of them the quality of life would probably skyrocket. A few thieves go a long way, and inspire the next generation to do the same.
And today's thief is tomorrow's mugger or murderer. Throw the book at them before somebody gets killed or hurt.
This whole philosophy is what made Guiliani popular in NYC - he cleaned the place up by spending money on petty crimes, and got rid of many of the big expensive crimes in the process.
The main issue is with linking GPL and non-GPL code. It is leagal to do so, but it is illegal to redistribute the result code under any license other than the GPL. Loading a kernel module involves linking it with the kernel.
Others have pointed out the reason for this far better than I could in a few sentences.
Well, you get what you pay for. The linux devs didn't charge you a cent for your OS, and haven't designed it to support your intended use. You aren't contributing anything back to the community, so the community doesn't really owe you anything. The linux kernel was designed to intentionally make what you're doing difficult - mainly to encourage your boss to get sick and tired of spending so much time on code updates and donate the code back. Your boss obviously hasn't spent quite enough money yet, perhaps one day they will...:)
If you want somebody who will let you have it your way, try buying an embedded OS, another unix, or maybe windows. But it won't be free-as-in-beer.
Actually, the parent's idea solves both. The terminal never sees the PIN, so it can't record it. The PIN is entered on the card itself, which would have a keyboard and LCD display. Nothing too fancy - think $5 calculator with crypto chip inside. Even if the PIN were captured via camera the combination of account number and PIN is not sufficient to access the account - you need the card's private key which never leaves the crypto chip.
I had a similar idea - but it would work offline as well as online. Merchant submits billing-request data packet containing merchant ID, transaction GUID, and amount. Card displays amount and user punches in PIN. Card timestamps and signs billing-request and gives the merchant the signed request and certificate. Merchant validates signature and submits packet to bank at next convenient opportunity. Private key and PIN never leave the card, and even if the PIN was videoed the card is still needed to authenticate transactions.
The card's private key would be generated internally and never leave. It would generate a standard SSL CSR and the bank would return a certificate.
I'd give the card a proximity interface, a USB interface, and if possible an acoustic modem. That covers all standard transaction modes.
I've been wanting this for a while, and I've yet to see an attack against it.
There are companies that make secure smartcards with a JVM inside for only $50 each. I'd think that if mass-deployed the cost would be far less to a major bank than the cost of fraud.
Slashdot Mirror
Uh, if you wanted to go with the flow, why aren't you using Windows?
It is plodding along just fine as well, with it's only-slightly-restrictive licensing practices.
Licenses matter. Linux will always do fine on the GPLv2 - companies like Tivo will ensure it! Now, whether the rest of the free software world ends up leaving it behind is a different matter. It obviously isn't going anywhere anytime soon, but stranger things have happened.
Well, to the average layman most of Einstein's theoretical work seems meaningless. But, when you build a photocopier anybody can see that there is something to it.
Ditto for everything else that physics has discovered. The value of the discoveries is appreciated when it is seen how these discoveries apply to the real world.
The issue with string theory is that while it is self-consistent, it seems like nobody is able to actually do anything useful with it, and to me that makes it an inadequate theory, because the proof is in the ability to apply the theory.
I can plot my movements for the entire day and fit them to a 47-degree polynomial with a decent level of error, and then wax philisophical about the general theory of human locomotion. And that would last about as long as it takes somebody to realize that five minutes after I publish the theory fails to account for my subsequent activities.
Given a complex enough equation you can fit any set of data. And given enough time you can even make that equation look "beautiful". What I want to know is how well it holds up six months from now without constant tweaking...
Actually, most foreign wars work like that. The British lost very few engagements in the US Revolutionary war. However, the war functioned to create lots of dead bodies in red uniforms, and that was not popular back home. The democratic people of Britain were scratching their heads over why exactly they were sending their soldiers overseas to kill a bunch of Americans who really just seemed to want to just be left alone.
I tend to agree that on a tactical and strategic level the war in Vietnam was successfully fought (not wisely fought, mind you, but even if we did manage to kill thousands of our own soldiers with dumb policies we still managed to get the job done on the battlefield in spite of ourselves). Now, the whole notion of limited war was dumb, and prevented the US from just cleaning up. Korea is a better example of what could have been achieved, but I'd hardly consider North Korea a great success story. And that brings up the bigger issue - if you want to get involved in foreign civil wars you're going to find that social change is a lot messier than just winning battles.
So, yes, in Vietnam the US probably didn't lose a single engagement, unless you count the decision to send troops in at all...
Yeah, and you'll be told to stop telling them how to run their lives.
Ask poor people to pay for drugs and you'll have a riot.
Offer them the opportunity to pay for other's drugs by selling worthless tickets and they'll line up with their paychecks...
But if you captured it using a framegrabber it would be:
1920*1080*30fps*12bpp*255min*60s/min/8bytes/bit = 1.4E12 bytes - and that isn't including the audio.
Even compressed HD uses quite a bit of space, but my point was that capturing DVI wasn't practical at all, even if you could afford the hardware (unless you buy realtime compression hardware).
Yeah, if you have a few TB of disk space. That card doesn't compress the signal - so you're drinking from a firehose. A 1 hour HD program would consume about 300GB of disk space if you used that card. Sure, you could then compress it down to a few GB, but not in realtime (figure 2 hours processing time for each 1 hour of recording time).
Want dual-tuners? Better have a bank of 6+ drives! Not only do you have to buffer all that video until it gets compressed, but you also have to stream it all at an insane rate of about 1.5Gbps - your SouthBridge won't even handle that, so you'd need a fancy motherboard design or several computers.
Consumer HD recording from DVI will only be practical when hardware mpeg2/4 compression becomes affordable. Right now the equipment for that is rack-mounted and costs a very pretty penny (computers, DSPs, signal-processors, probably lots of fiber, etc). TV studios use it, and they have trouble getting it all right...
Just to add a little to the other reply - your mythtv stream is only a few megabits/sec. Certainly not 700Mb/s! You might be able to record at that rate with the right drive setup, but you'd fill your hard drive mighty fast (80 megabytes per second eats through space pretty quick).
Go ahead and check the mythtv-users list - this is a common topic. The hardware capable of compressing live HD is very expensive - studio gear. We're not talking Apollo-mission cutting edge, but even the TV studios have difficulty with live HD streams (I know somebody who works in the industry).
Anybody recording HD using myth is recording compressed MPEG2 - not DVI. In fact, there is a company that will mod your cable box with DVI/HDMI-only output to add a firewire port so that you can record the HD stream at a decent rate.
Well, if person on the top-500 list has two kids, and they get every dime of the cash (no taxes, etc), neither would probably end up on the top-500 list.
Problems of economic division don't really hinge on the top-500 list - we're only talking 500 people here. And I do think that inheritance does reduce the incentive to achieve, and reduces the effective spending of cash.
My feeling is that inheritances below $1 million per recipient should be moderately taxed, and anything above that should be taxed at around the 90% level. If you can't get started in life with $1 mil cash then you'll be bound for welfare no matter how much you're given. Go ahead and adjust the exemption for COL and all that. Even with a 90% tax Bill Gates's kids would still be billionares.
If somebody earns their pay they shouldn't be punished for their success. I'm OK with progressive taxation, but let's not get to crazy with that. Unearned income is a different story (although investments shouldn't be too heavily taxed or nobody would invest).
Yet, in a perfect system, the benefit would be proportional to productivity -- which isn't close to reality.
Pick a person at random. Determine how much THEY increased their productivity, and look at how much their wages have increased. It will probably be decently correllated.
Now, if a person's employer bought a fancy machine that made their employees more productive, then the employees have NOT increased their own productivity, and consequently they will see very little of the productivity increase in their income. People raise their OWN productivity by obtaining an education, inventing better processes, etc. You don't raise your OWN productivity by showing up to work and doing what you're told, and people who do this will never make the kind of money those who do otherwise earn.
Those who contribute the most to an enterprise are going to reap the rewards. Sure, there is some random variation - some people inherit money (I support estate taxes for this reason), and some people win the lottery (a good way to tax the poor without making them upset). For the most part, though, if you want to get ahead in society you have to do more than show up.
Uh, why would I or somebody else want to take a job that pays $8/hour? You could probably do better than that running an open-source project and soliciting donations, let alone something that is actually useful to society.
If you're making $8/hour, I'd recommend getting an education and finding something to do with your life. You can do better than that trading on ebay. And a minimal education doesn't have to cost money - read wikipedia or something and you'll figure out something useful to do with your life.
One indicator that's often used to assess poverty is the percentage of income that goes into housing: if you're paying 50% of your income just to keep your mortgage steady, you're worse off than someone who pays 40% of their income to a landlord instead of to the bank.
Uh, sure, but be careful about how you measure it.
Money paid to mortgage-INTEREST is a housing expense. Money paid to mortgage-PRINCIPAL is not - you're paying it to yourself.
If you're just talking size of payments, I'd much rather pay 50% of my income to a mortgage than 40% to rent for 30 years! When I'm done in 30 years I have a nice house to my name, and I can sell it for a good chunk of cash. If I rent I own nothing, and I'm still paying 40% of my income to continue to rent.
Buying isn't always better than renting, but most of the time it is. How do you think the landlords make their money?
As far as air conditioning / etc being cheap goes - it isn't so much the cost to buy the unit as the cost to operate it. If you can afford those kinds of electric bills then you aren't actually in horribly bad shape.
No matter how you slice it, poverty isn't a bucket of roses.
And good thing that! Otherwise everybody would just sit around being poor!
Relax - no matter what the plans are you'll be able to buy one for 1/5th of cost. Just search ebay a month after they come out. Kids in developed countries will "lose" them and turn a nice profit (since they were free to them to begin with).
Why they're looking to promote the gray market by not selling retail is beyond me...
Agreed. For all the cost of these 10,000 year storage facilities and all the argument that goes on, they could just station 5000 marines at a number of reactors and just sell the fuel to operators at subsidized rates (or charge a pretty penny for disposal of non-reclaimed waste). If you only reburn the waste at a half-dozen facilities and run them as highly secure installations then it is very unlikely terrorists would get in. Honestly, rent-a-cops are enough to deter most terrorists as long as you have locked doors and such to slow down entry/escape from a secured facility. But hey, we're talking nukes, so go ahead and spend a little more to put black copters and tanks outside - it has to be cheaper than Yucca Mountain.
Well, if the only concern was DoS or other non-customer-approved use there is a simple solution - just give customers a website allowing them to specify domains/netblocks that are exempt from QoS filters. If they can base their filters on who is paying them, they can certainly base it on what the customer wants.
Implement QoS in a way that any given customer on average gets the same bandwidth, but QoS bits can be used to prioritize traffic for a single customer, or to prioritize across multiple customers for short bursts only. Also allow end-users to buy higher levels of service (which is already something ISPs sell).
Make the end-user the customer - not the remote site.
You can avoid congestion the same way you've always done it: don't do too many things at once over the same connection.
Well, only if the congestion is at your DSL modem/etc. If the congestion is at the OC3 line leading out of your ISP you'll have issues no matter what you do. Right now ISPs avoid such congestion, since customers would complain. Once they start charging for QoS they'll be sure to stop expanding their bandwidth - otherwise they don't have anything to sell (who needs QoS when you aren't bandwidth-starved?). Then pay-to-play websites will get the same service they get now, and anybody else will get degraded service. If you are in an area like mine you'll only have one or two choices for an ISP, and both will probably be doing the same thing.
If the telecoms really just cared about congestion at your DSL modem they'd just stop filtering out the QoS bits and instead ignore them everywhere but on the two ends of your DSL line. You'd get all the benefits of QoS without any disruption to your neighbors - if this remains the main choke-point. No need to charge money for it.
Then why don't you do your neighbors a favor and try to start your own cable company. You'll see how well the friendly monopoly treats you... :)
If it were purely tradition there would be SOME town SOMEWHERE that has a couple of utility providers. The fact that this is unheard of indicates that this is likely a natural monopoly.
About the only areas you see competition with local utilities is stuff like WiFi and satellite-cable. And this is precisely because they get rid of the huge localized investments that make it impossible to compete with utilities. If such a technology existed for broadband I'd be less in favor of net neutrality. I've seen Verizon's idea of premium internet service - I once tried it out on my cell phone. 95% of the WAP sites out there weren't reachable - because they didn't pay to play.
I'd be happy if linux worked right with my mouse when I use my cheap KVM. It works fine with windows, but not with linux. Sure, I understand the issues with mice being stateless and all that, but obviously MS has a fix. On my desk I have one keyboard, one monitor, and two mice as a result. And no, I don't want to go out and spend $300 on a nicer unit...
But it wouldn't be proprietary if you gave it to them...
Well, there isn't much point if they're going to keep it to themselves. How many hobbyists need to hang onto the code, and wouldn't donate it back to the kernel team?
But only the closed-source vendors would be able to maintain it. Suppose I have a TNT2 card - do you think that Nvidia is going to release a driver for it when linux 2.8 comes out?
The idea is to get rid of the closed-source easy-out - which should result in more code being opened. Sure, some code just won't get used at all, but that puts vendors who refuse to open their code at a disadvantage, at least among linux users. And why should Nvidia profit by selling cards to linux users when they don't release their code? The drivers wouldn't work without the kernel, and the kernel was made GPL to force redistributers to release their sources. Now, the kernel won't go GPLv3 anytime soon, but other software will, and future GPL versions will only tighten some of the closed-source loopholes that currently exist.
It has worked pretty well so far. And if somebody wants to make their own middle-layer they can, just don't expect the kernel team to help you out.
First - I'm very sorry to hear about this tragedy - I honestly have difficult replying just because I can't really equate the seriousness of what befell your mother with the content of this post.
That said, my feeling is that the only reason the police even investigated this theft is due to the associated near-murder of your mother. As many others have indicated, if a crime is "just a theft" it often gets ignored.
The tragedy is that most likely the thug that nearly killed your mother probably stole something before then, and maybe in one of those cases they could have been tracked down (maybe they made a mistake). If they were punished for the minor crime of theft then maybe things would have turned out better for your mother.
In my opinion crimes like theft and vandalism should be treated very seriously - they are personal in nature and if somebody is willing to smash your windows they will eventually be willing to smash you. Even if only $100 is stolen the police should be willing to spend a good sum to apprehend the criminal - eventually it will pay off.
Ok, but what if the owner wasn't right there?
Or what if the guy was attacked by a mountain lion?
Or what if the guy hit a dog with a branch to fend it off and the owner went nuts and attacked him without a gun? He may have gotten 10 years but at least he is alive..
It is all about escalation and quality of life. When I was in school somebody spray-painted the outside wall of the school. The next morning the principal delivered a long sermon, and within about 24 hours the pain was sandblasted off - probably at a cost of $10k or more. However, the move was effective - no more sparypainting for as long as I was there. If the paint were left up I'm sure somebody would have added to it within a week. Sure, the paint doesn't cost anything, but it makes th school look like a ghetto and probably would tend to result in the students acting more like they lived in one.
If somebody steals a $250 stereo the police should spend $40k tracking the guy down. Then fine and/or imprison the criminal for the full cost of the investigation - if they're underage put a lean on the parent's house. A town with a theft problem probably has only 10 thieves in it, and if you got rid of them the quality of life would probably skyrocket. A few thieves go a long way, and inspire the next generation to do the same.
And today's thief is tomorrow's mugger or murderer. Throw the book at them before somebody gets killed or hurt.
This whole philosophy is what made Guiliani popular in NYC - he cleaned the place up by spending money on petty crimes, and got rid of many of the big expensive crimes in the process.
Now, the war on drugs is a different story...
The main issue is with linking GPL and non-GPL code. It is leagal to do so, but it is illegal to redistribute the result code under any license other than the GPL. Loading a kernel module involves linking it with the kernel.
Others have pointed out the reason for this far better than I could in a few sentences.
Well, you get what you pay for. The linux devs didn't charge you a cent for your OS, and haven't designed it to support your intended use. You aren't contributing anything back to the community, so the community doesn't really owe you anything. The linux kernel was designed to intentionally make what you're doing difficult - mainly to encourage your boss to get sick and tired of spending so much time on code updates and donate the code back. Your boss obviously hasn't spent quite enough money yet, perhaps one day they will... :)
If you want somebody who will let you have it your way, try buying an embedded OS, another unix, or maybe windows. But it won't be free-as-in-beer.
Sorry, "see" as in "visualize" or "conceive". :)
Actually, the parent's idea solves both. The terminal never sees the PIN, so it can't record it. The PIN is entered on the card itself, which would have a keyboard and LCD display. Nothing too fancy - think $5 calculator with crypto chip inside. Even if the PIN were captured via camera the combination of account number and PIN is not sufficient to access the account - you need the card's private key which never leaves the crypto chip.
I had a similar idea - but it would work offline as well as online. Merchant submits billing-request data packet containing merchant ID, transaction GUID, and amount. Card displays amount and user punches in PIN. Card timestamps and signs billing-request and gives the merchant the signed request and certificate. Merchant validates signature and submits packet to bank at next convenient opportunity. Private key and PIN never leave the card, and even if the PIN was videoed the card is still needed to authenticate transactions.
The card's private key would be generated internally and never leave. It would generate a standard SSL CSR and the bank would return a certificate.
I'd give the card a proximity interface, a USB interface, and if possible an acoustic modem. That covers all standard transaction modes.
I've been wanting this for a while, and I've yet to see an attack against it.
There are companies that make secure smartcards with a JVM inside for only $50 each. I'd think that if mass-deployed the cost would be far less to a major bank than the cost of fraud.