People think I'm paranoid because I encrypt all my drives... but when I lose a disk I never have to wonder if it potentially ended up in the wrong hands. Too bad it's only done by us loonies and not as standard practice everywhere.
That's actually what I was suggesting in the third paragraph. You don't want the motors in the wheels themselves if you can avoid it: unsprung weight hurts much more than weight on the body, so it's better to put the motors inboard and drive the wheels through halfshafts, but otherwise ripping out all the drivetrain you can is a win.
First, classical body-on-frame, or a unibody crossover? Next, do you want it to be 4WD when you're done?
The easy way is to replace the internal combustion motor with an electric one, preserving the existing driveshaft, transfer case, transmission, etc.
If you're up to the harder method though: rip out ALL of the existing drivetrain. If it's an independent suspension crossover you can even lose the axles and go with one motor per wheel. If it's a truck-style one you will probably need to keep at least the rear axle, but you could mount the motor right up to the differential, likely with a reduction gear. Do the same thing on the front if you want 4WD.
On a body-on-frame SUV, this might clear enough room between the frame rails to fill in battery packs. Unibody models might not be so lucky.
If you want it to be good 4WD, actively proportion the torque between axles or wheels (depending on 2 or 4 motors). If one axle / wheel goes significantly faster, reduce power and redirect it to the others. It won't be quite as capable as real diff lockers, but it won't get stuck nearly as easily as AWD (three open diffs) which can't keep going when one wheel loses grip.
Consider where you want your weight - more batteries or more engine. I would love to see this done with a "more batteries" approach - full electric running for 40-60 miles then running a small engine fairly hard for extended range, but which is never required for in-town use (IE, like a Volt), instead of a large engine and small batteries where the electric assist just improves the fuel efficiency (Prius). The batteries cost more, but it would make this conversion much more interesting to me.
I bought a $3 tube years ago and I still have plenty left. You're only supposed to use a dot the size of a BB. If you smear it all over the CPU you're doing it wrong - aside from the mess, that's guaranteed to create air pockets.
(The cheap stuff is fine too. The expensive stuff may conduct heat better, but the layer of goo is so thin that it's only a fairly small percentage of your heat resistance.)
Still, these pads are interesting. It looks reliable and less prone to noob mistakes. Too bad it's Sony.
It's secure against its intended threat: some guy you just arrested and frisked (and who therefore probably doesn't have a key). The slightly more complicated key isn't to prevent duplicates from being made. It's to prevent the arrestee from picking up a paperclip from the gutter and getting free.
I agree it's flawed, but not fundamentally so. In the Airbus model where there is heavy computer assistance, the pilots are certainly more disoriented when Things Go Wrong badly enough and the plane finally admits it doesn't know what it's doing and starts progressively giving more control over to the pilots (Alternate Law 1 and 2, Direct Law). However, with all the computer assistance, Things Go Wrong less frequently.
Airbus experiences more computer-failure crashes, whereas Boeing experiences more pilot-error crashes. The net safety record for both is about the same.
I think it's not fundamentally flawed because while there's a limit to how much you can improve pilot error (the job is pretty routine and boring 99.9% of the time, and while good training usually takes care of the 0.1% case, humans are simply going to have a limit to how quickly they can reorient in an emergency), the computers can be continuously improved as flaws are found.
Sometimes I do make flippant remark or make an attempt a humor that (rightfully) gets modded down
My pet peeve: sometimes I make some vaguely amusing remark in the middle of an otherwise well thought out post. Someone moderates it "Funny", but it really wasn't, nor was it meant to be. Then future mods look at it as a trainwrecked attempt at humor rather than being mis-moderated, and it gets pounded to the ground with "overrated". It's really frustrating to have that happen when I put a lot of effort into a long, well-researched comment.
I'm not sure what could be done about it, though. Perhaps hide the "Funny" flag from moderators to prevent the bias?
The short version: Instead of putting your data in the code, you put your code in the data, and that makes your code much easier to understand and reuse. Here's a quick Ruby example of doing something procedurally:
The idea is that by attaching your code to the data, the data becomes "smart" and knows how to process itself. This is a huge benefit for code reuse since the person reusing your classes doesn't have to understand how they represent data internally, or the mechanics of how they process the data. It's already slightly easier to understand with this short example, and the effect is magnified greatly as the size of the codebase grows.
Regardless of whether it's a good idea or not, I'm just talking about the way things are. 30-year careers are just unheard-of.
In other industries it's certainly different, but in software companies where there's not as much long-term infrastructure to support: build servers are simply replaced every few years, the WAN is just a standard bunch of VPNs between offices so anyone can figure it out once they have a copy of the network map. The product itself has long term needs and hanging onto the architects and top developers is important, but everywhere I've been, turnover is just customary.
Not that I particularly love Verizon, but their network is the last thing I'll criticize. They have LTE coverage (the spiffiest, shiniest 4G air interface available, ITU aside) in most major cities, and EV-DO (probably the best of the 3G standards) in pretty much any town with enough people to have an orgy. By most reviews, their coverage is much less spotty and they drop fewer calls than everyone else.
I won't argue if you want to talk about customer service, or pricing structures, or apparently the net neutrality stance of the organization. They have problems. But the network isn't one, especially compared to their competitors.
30 years? My whole career is in software companies and I have never worked at any company that even pretended that anyone was going to stick around that long. Retirement consists of a good 401k plan, not a pension. Commitment comes in the form of vesting schedules. I have achieved seniority in several departments by sticking around 3-4 years. The very idea of working for one place for 30 years straight just boggles my mind, and honestly it's not an attractive idea to me at all.
They are telling the power company how much electricity you are using.
I don't care if the power company knows. What I care about is who they're selling that information to in minute-by-minute increments, which therefore includes what hours I keep, when I'm watching TV, etc. All they have to do is create a reasonable privacy policy.
The RF stuff is 1% tinfoil hatters and 99% red herring to make the anti-smart-meter crowd look like a bunch of tinfoil hatters.
100 KW thruster, assuming pretty near 100% efficiency - about 25 pounds thrust.
Can you show your work on that? I did some questionable math and came up with ~5000N. Reading the hall effect thruster article suggests 1.35kW produces 83 mN. That certainly sounds like you're much closer than I, but I'd just like to see how you got there. If it's true, you're right: that's a completely unreasonable for transfers, and would only be useful for minor corrections to existing trajectories.
As I noted below, hall effect thrusters are an odd choice for this kind of mission, but no, it wouldn't be months. They're talking about 100KW thrusters. Those would be able to get you there in days to weeks, not months. Still, I don't see the logic in them. You lose a lot of the thrusters' specific impulse efficiency by having to use a crap transfer trajectory.
The optimal transfer injection burn is a short, very strong impulse, not a long gentle one. A worst case (continuous burn all the way there, just changing the vector as you go) requires > 150% more delta-v (which comes out to more than 2.5 x the propellant - the extra weight holds you back at the worst time, when you're deep in Earth's gravity well) than the optimal case (a brief kick to get out of Earth orbit, and another brief kick when you arrive at the moon, or back at Earth if you take a free return trajectory). The reason is delta-v is much more efficient when applied at perigee than apogee - you spend much less time and energy fighting gravity.
For those unfamiliar with the tradeoffs: Hall effect thrusters make fairly efficient use of the reaction mass - about 2000s, compared to ~250 for solid rockets or ~300-400 for liquid rockets. That means a considerable increase in your delta-v - since you only need 10-20% as much reaction mass for the same impulse, you get 5-10x more delta-v. Great, right?
The trouble is that you need a power source. Liquid fuel rockets just burn the propellant. Hall effect thrusters (and other ion thrusters) need a power source in addition to the propellant.
This is a great tradeoff for stationkeeping on satellites - you only need tiny amounts of thrust, so you can easily generate enough power using solar cells or a RTG. Thus the very efficient use of reaction mass means a much longer useful life, or more useful payload in your satellite for a given launch mass, etc. It's just plain more efficient.
But this isn't like that. They seem to want to use them to perform the Hohmann tranfer. That means having a very high thrust for a short duration - not just because you want to get there more quickly, but because it's much more efficient than a long continuous burn.
They're talking about 100KW. That seems low. Ballpark 5000 newtons of thrust... Compare to the Apollo command/service module at ~90,000 newtons. Thus they'll need a fairly long burn at that power. How the heck are they generate that kind of power for a long duration?
We see a rash of ISP's fighting copyright shakedown attempts by doing what Sonic is now doing... However, it is still about costs. Sonic just thinks they've hit upon a quick and cheep way to deal with this issue.
Sonic has been standing up for their users for a long time. They're not your usual money-grubbing operation. Dane built the thing from the ground up and still runs it with the same ideals he started with. From being a customer for a long time, to knowing the people that work there (Dane and Scott were part of the community before they jumped into the ISP business), to watching them take the high ground numerous times... It's not to make a quick buck, but to really earn it by doing the right thing. I know that's not business as usual in the US, but occasionally a truly good company like this crops up.
In which case, the seller is the one paying the unwanted middleman...
I assume you mean buyer. No, the buyer doesn't pay for it either - they get a better deal than they would have otherwise.
Let's say we have market A: Bid: $10, Ask $11. And market B: Bid $12, Ask $13.
A buyer comes to market B and places a market order for one unit (not a bid, which may or not be filled; a market order means "buy immediately at the best available asking price").
With no arbitrage: The buyer on B pays $13, the seller on B gets $13, the seller on A doesn't make a sale.
With arbitrage: Some HFT buys 1 unit of commodity on A for $11, then immediately sells it to the buyer on B for $12.50. (Technically they buy and sell different units of the commodity, since the commodity can't be immediately transferred.) The seller on A makes a sale, the seller on B doesn't.
The guy who loses out is the seller on B who fails to earn his higher asking price. Are we supposed to feel sorry for his lost sale when a better deal was available? The HFT acted as a buying service to grab that deal on behalf of the buyer. The buyer couldn't take advantage the deal on A directly because their currency was deposited with an agent on exchange B. Likewise the seller on A couldn't take direct advantage of the conditions on market B. Both of them benefit, and the HFT makes the difference at the cost of having their funds tied up until the transaction settles.
The end result is that the prices on markets A and B end up pulled closer together (as close as the HFT guys can get while still making money). That means more even and fair pricing for everyone on both markets.
Slashdot Mirror
People think I'm paranoid because I encrypt all my drives... but when I lose a disk I never have to wonder if it potentially ended up in the wrong hands. Too bad it's only done by us loonies and not as standard practice everywhere.
That's actually what I was suggesting in the third paragraph. You don't want the motors in the wheels themselves if you can avoid it: unsprung weight hurts much more than weight on the body, so it's better to put the motors inboard and drive the wheels through halfshafts, but otherwise ripping out all the drivetrain you can is a win.
First, classical body-on-frame, or a unibody crossover? Next, do you want it to be 4WD when you're done?
The easy way is to replace the internal combustion motor with an electric one, preserving the existing driveshaft, transfer case, transmission, etc.
If you're up to the harder method though: rip out ALL of the existing drivetrain. If it's an independent suspension crossover you can even lose the axles and go with one motor per wheel. If it's a truck-style one you will probably need to keep at least the rear axle, but you could mount the motor right up to the differential, likely with a reduction gear. Do the same thing on the front if you want 4WD.
On a body-on-frame SUV, this might clear enough room between the frame rails to fill in battery packs. Unibody models might not be so lucky.
If you want it to be good 4WD, actively proportion the torque between axles or wheels (depending on 2 or 4 motors). If one axle / wheel goes significantly faster, reduce power and redirect it to the others. It won't be quite as capable as real diff lockers, but it won't get stuck nearly as easily as AWD (three open diffs) which can't keep going when one wheel loses grip.
Consider where you want your weight - more batteries or more engine. I would love to see this done with a "more batteries" approach - full electric running for 40-60 miles then running a small engine fairly hard for extended range, but which is never required for in-town use (IE, like a Volt), instead of a large engine and small batteries where the electric assist just improves the fuel efficiency (Prius). The batteries cost more, but it would make this conversion much more interesting to me.
I bought a $3 tube years ago and I still have plenty left. You're only supposed to use a dot the size of a BB. If you smear it all over the CPU you're doing it wrong - aside from the mess, that's guaranteed to create air pockets.
(The cheap stuff is fine too. The expensive stuff may conduct heat better, but the layer of goo is so thin that it's only a fairly small percentage of your heat resistance.)
Still, these pads are interesting. It looks reliable and less prone to noob mistakes. Too bad it's Sony.
It's secure against its intended threat: some guy you just arrested and frisked (and who therefore probably doesn't have a key). The slightly more complicated key isn't to prevent duplicates from being made. It's to prevent the arrestee from picking up a paperclip from the gutter and getting free.
He is at least an astute businessman rather than a hack. The bar is pretty low.
I agree it's flawed, but not fundamentally so. In the Airbus model where there is heavy computer assistance, the pilots are certainly more disoriented when Things Go Wrong badly enough and the plane finally admits it doesn't know what it's doing and starts progressively giving more control over to the pilots (Alternate Law 1 and 2, Direct Law). However, with all the computer assistance, Things Go Wrong less frequently.
Airbus experiences more computer-failure crashes, whereas Boeing experiences more pilot-error crashes. The net safety record for both is about the same.
I think it's not fundamentally flawed because while there's a limit to how much you can improve pilot error (the job is pretty routine and boring 99.9% of the time, and while good training usually takes care of the 0.1% case, humans are simply going to have a limit to how quickly they can reorient in an emergency), the computers can be continuously improved as flaws are found.
Regardless of how you feel about Warren Buffett, he'll be better for us and better for the newspaper industry than Rupert Murdoch.
This already exists. It's called a taxi. I agree that there's an advantage in making taxis more efficient and lower cost though.
Oh FFS. :)
Sometimes I do make flippant remark or make an attempt a humor that (rightfully) gets modded down
My pet peeve: sometimes I make some vaguely amusing remark in the middle of an otherwise well thought out post. Someone moderates it "Funny", but it really wasn't, nor was it meant to be. Then future mods look at it as a trainwrecked attempt at humor rather than being mis-moderated, and it gets pounded to the ground with "overrated". It's really frustrating to have that happen when I put a lot of effort into a long, well-researched comment.
I'm not sure what could be done about it, though. Perhaps hide the "Funny" flag from moderators to prevent the bias?
what is the big draw to OO? The killer feature?
The short version: Instead of putting your data in the code, you put your code in the data, and that makes your code much easier to understand and reuse. Here's a quick Ruby example of doing something procedurally:
def degCtoF(c) ; return (c * 9.0 / 5.0 ) + 32 ; end
def degFtoC(f) ; return (f - 32) * 5.0/9.0 ; end
def format_c_temp(c) ; return "%s degrees F, %s degress C" % [degCtoF(c), c] ; end
oven_temperature_C = 55
puts format_c_temp(oven_temperature_C)
And then again doing it OO:
class Temperature :degC
attr_accessor
def degF ; return (@degC * 9.0 / 5.0) + 32 ; end
def to_s ; return "%s degrees F, %s degrees C" % [degF, degC] ; end
end
oven_temperature = Temperature.new
oven_temperature.degC = 55
puts oven_temperature.to_s
The idea is that by attaching your code to the data, the data becomes "smart" and knows how to process itself. This is a huge benefit for code reuse since the person reusing your classes doesn't have to understand how they represent data internally, or the mechanics of how they process the data. It's already slightly easier to understand with this short example, and the effect is magnified greatly as the size of the codebase grows.
on any computer with java
Fail.
No. Compared to the OP which is a closed-source, native-binary-only solution: Win.
People who think twice about clicking this link generally aren't affected by dnschanger in the first place.
... I use LastPass for everything. I keep the Lastpass password / encryption key on paper in a secure location where it can be recovered by my SO.
The GP and I were talking about cellular data, not FIOS.
Regardless of whether it's a good idea or not, I'm just talking about the way things are. 30-year careers are just unheard-of.
In other industries it's certainly different, but in software companies where there's not as much long-term infrastructure to support: build servers are simply replaced every few years, the WAN is just a standard bunch of VPNs between offices so anyone can figure it out once they have a copy of the network map. The product itself has long term needs and hanging onto the architects and top developers is important, but everywhere I've been, turnover is just customary.
Not that I particularly love Verizon, but their network is the last thing I'll criticize. They have LTE coverage (the spiffiest, shiniest 4G air interface available, ITU aside) in most major cities, and EV-DO (probably the best of the 3G standards) in pretty much any town with enough people to have an orgy. By most reviews, their coverage is much less spotty and they drop fewer calls than everyone else.
I won't argue if you want to talk about customer service, or pricing structures, or apparently the net neutrality stance of the organization. They have problems. But the network isn't one, especially compared to their competitors.
30 years? My whole career is in software companies and I have never worked at any company that even pretended that anyone was going to stick around that long. Retirement consists of a good 401k plan, not a pension. Commitment comes in the form of vesting schedules. I have achieved seniority in several departments by sticking around 3-4 years. The very idea of working for one place for 30 years straight just boggles my mind, and honestly it's not an attractive idea to me at all.
They are telling the power company how much electricity you are using.
I don't care if the power company knows. What I care about is who they're selling that information to in minute-by-minute increments, which therefore includes what hours I keep, when I'm watching TV, etc. All they have to do is create a reasonable privacy policy.
The RF stuff is 1% tinfoil hatters and 99% red herring to make the anti-smart-meter crowd look like a bunch of tinfoil hatters.
100 KW thruster, assuming pretty near 100% efficiency - about 25 pounds thrust.
Can you show your work on that? I did some questionable math and came up with ~5000N. Reading the hall effect thruster article suggests 1.35kW produces 83 mN. That certainly sounds like you're much closer than I, but I'd just like to see how you got there. If it's true, you're right: that's a completely unreasonable for transfers, and would only be useful for minor corrections to existing trajectories.
As I noted below, hall effect thrusters are an odd choice for this kind of mission, but no, it wouldn't be months. They're talking about 100KW thrusters. Those would be able to get you there in days to weeks, not months. Still, I don't see the logic in them. You lose a lot of the thrusters' specific impulse efficiency by having to use a crap transfer trajectory.
The optimal transfer injection burn is a short, very strong impulse, not a long gentle one. A worst case (continuous burn all the way there, just changing the vector as you go) requires > 150% more delta-v (which comes out to more than 2.5 x the propellant - the extra weight holds you back at the worst time, when you're deep in Earth's gravity well) than the optimal case (a brief kick to get out of Earth orbit, and another brief kick when you arrive at the moon, or back at Earth if you take a free return trajectory). The reason is delta-v is much more efficient when applied at perigee than apogee - you spend much less time and energy fighting gravity.
For those unfamiliar with the tradeoffs: Hall effect thrusters make fairly efficient use of the reaction mass - about 2000s, compared to ~250 for solid rockets or ~300-400 for liquid rockets. That means a considerable increase in your delta-v - since you only need 10-20% as much reaction mass for the same impulse, you get 5-10x more delta-v. Great, right?
The trouble is that you need a power source. Liquid fuel rockets just burn the propellant. Hall effect thrusters (and other ion thrusters) need a power source in addition to the propellant.
This is a great tradeoff for stationkeeping on satellites - you only need tiny amounts of thrust, so you can easily generate enough power using solar cells or a RTG. Thus the very efficient use of reaction mass means a much longer useful life, or more useful payload in your satellite for a given launch mass, etc. It's just plain more efficient.
But this isn't like that. They seem to want to use them to perform the Hohmann tranfer. That means having a very high thrust for a short duration - not just because you want to get there more quickly, but because it's much more efficient than a long continuous burn.
They're talking about 100KW. That seems low. Ballpark 5000 newtons of thrust... Compare to the Apollo command/service module at ~90,000 newtons. Thus they'll need a fairly long burn at that power. How the heck are they generate that kind of power for a long duration?
We see a rash of ISP's fighting copyright shakedown attempts by doing what Sonic is now doing ... However, it is still about costs. Sonic just thinks they've hit upon a quick and cheep way to deal with this issue.
Sonic has been standing up for their users for a long time. They're not your usual money-grubbing operation. Dane built the thing from the ground up and still runs it with the same ideals he started with. From being a customer for a long time, to knowing the people that work there (Dane and Scott were part of the community before they jumped into the ISP business), to watching them take the high ground numerous times... It's not to make a quick buck, but to really earn it by doing the right thing. I know that's not business as usual in the US, but occasionally a truly good company like this crops up.
In which case, the seller is the one paying the unwanted middleman ...
I assume you mean buyer. No, the buyer doesn't pay for it either - they get a better deal than they would have otherwise.
Let's say we have market A: Bid: $10, Ask $11.
And market B: Bid $12, Ask $13.
A buyer comes to market B and places a market order for one unit (not a bid, which may or not be filled; a market order means "buy immediately at the best available asking price").
With no arbitrage: The buyer on B pays $13, the seller on B gets $13, the seller on A doesn't make a sale.
With arbitrage: Some HFT buys 1 unit of commodity on A for $11, then immediately sells it to the buyer on B for $12.50. (Technically they buy and sell different units of the commodity, since the commodity can't be immediately transferred.) The seller on A makes a sale, the seller on B doesn't.
The guy who loses out is the seller on B who fails to earn his higher asking price. Are we supposed to feel sorry for his lost sale when a better deal was available? The HFT acted as a buying service to grab that deal on behalf of the buyer. The buyer couldn't take advantage the deal on A directly because their currency was deposited with an agent on exchange B. Likewise the seller on A couldn't take direct advantage of the conditions on market B. Both of them benefit, and the HFT makes the difference at the cost of having their funds tied up until the transaction settles.
The end result is that the prices on markets A and B end up pulled closer together (as close as the HFT guys can get while still making money). That means more even and fair pricing for everyone on both markets.