Yup - clinical trials cost a fortune. Most of the costs of developing a drug are in the trials - tens of millions of dollars each and hundreds of millions of dollars in total (and that is when they're run assembly-line style by organizations that do nothing but run and supply trials 24x7).
I think half of it is essentially corporate welfare for doctors and hospitals. The costs you point out are very real of course. You also need to factor in that if you pay a doctor $100 per visit per subject and somebody else offers $150 per visit per subject then you won't get any subjects. There seems to be a lot more demand for subjects than supply. Of course, the actual volunteers don't see a dime of this money, and chances are their doctor doesn't mention to them that they are recommending experimental treatment A over experimental treatment B because they get more money out of it. Not that doctors would actually do anything like base treatment decisions on their financial interests...
Actually, if you look at the FDA's debarment list (a list of people forbidden to work on pharmaceuticals or trials), you'll find most are doctors. The most typical offense is signing up people for trials who shouldn't be in the trials so that they can get paid all the associated fees. This is bad for everybody. The patient gets experimental drugs that may not help them, or might even hurt them. The pharmaceutical company gets clinical trial data that contains more noise, which means they have to subject even more people to experimental medication to find out if it works or not, and pay more money as a result. The public ends up with less accurate clinical trial data, which might mean that a drug that actually works doesn't get approved (which means people don't benefit from it, or maybe they keep taking inferior drugs that are actually less safe even though the trial data suggests otherwise).
Human clinical trials are a mess. Their ability to generate meaningful data is also very weak compared to things like animal testing. The problem is that they're all we have, so we need to make the most of them.
The GP is right that the only way we can see independent trials happen is if the government funds them - or maybe VERY LARGE charities. I'm all for it - if an independent trial might benefit the public by showing benefit from a cheap "new" drug, or nail the coffin in a bad drug, then let's have it. And I'm generally in favor of drug patents. Competition is always good for consumers and it keeps private companies more honest.
Also - if cars drove themselves we wouldn't need so many of them. My car could drop me off at work, and then drive home to take care of my wife all day until I need it again.
If I needed a car unexpectedly I'd just punch something into my smartphone and in 2 mins a rental/taxi would drive up to pick me up.
Public transit would make a whole lot more sense if a car dropped me off on the train platform, and another was sitting there ready to pick me up for the last mile when I got off at the other end.
For a big family vacation I could have a second car just shuttle my luggage. Maybe it would drive at 45mph for maximum economy and leave early or arrive late.
No need to own a big truck, but I could have one with two minutes notice if I need to haul something. If I need to drop something off at a neighbor's there would be no need for me to even ride along for the trip as long as they're to unload at the destination.
Stores wouldn't need parking lots. Each town would have big parking garages strategically located, and cars would drop people off and go find something to do with themselves until needed. Just think about how much less asphault we'd need, and how much that would be better for the enviornment/etc.
Oh, no cars idling anywhere either, no red lights, etc. That has to be great for improving gas mileage.
Lots of options if you get rid of the human drivers. For far less than we spend on all the problems caused by our current transportation system we could probably fix it...
I'm not sure we're just talking about education - we're talking about reaction times and fallibility.
Why shouldn't my car travel an average speed of less than 45mph on minor roads without stops on my way to work? Right now I probably average 20mph tops - 45mph when driving, and 0mph at lights/etc.
Automated vehicles would just proceed at fairly high speeds the entire distance of a trip with greatly reduced risk of accidents/etc. On highways they would cruise at 100mph or so, with a few inches clearance between cars to maximize aerodynamics. Presumably cars following the same routes would huddle together, ordering themselves to minimize aerodyanmic profile.
You can't mix this kind of system with human drivers, regardless of how much training they have. What will you do when you come to your first intersection between two four-lane roads without a traffic light with traffic moving through at highway speeds in all directions simultaneously? With automated cars each car would get a reservation for every spot along the intersection, with sufficient margins for error/failures/etc. No human could safely navigate under these conditions.
The real issues (as others have pointed out) are the liabilities. Today Ford's cars kill maybe 1k people per year, but they have no liability. If they automated their cards and killed only two people per year they'd be sued for everything they own. Our legal system doesn't care that they've saved 998 people per year - only that they are now subject to some kind of liability for the 2 they didn't manage to save.
Utilities should be BORING industries. By all means they should improve bandwidth/etc, and be able to sell that. However, they should be pipes and wires.
The problem is all the vertical integration. Nobody wants to be a company that guarantees shareholders 25cents per share every quarter from now until eternity. They want to be able to promise double-digit earnings growth, and that requires the ability to grow markets. However, when your market is ever home in a 10 mile radius who could own a phone, and they already own phones, then there is no room to grow.
The problem is greed. It isn't like the CEO of a boring utility company doesn't make a good six-figure salary. However, who wants 6-figures when they could be the next Bill Gates? Well, if you want to leave and start your own company that's fine, but when you want to use a government-granted monopoly as the springboard for world conquest then don't be surprised when taxpayers start complaining.
Yup - we should get the government to handle this because they have such a better track record in managing unusual risks, like they did with the toxic asset crash and the space shuttle.
Don't get me wrong - we do need more oversight, and we do need to make companies cover their externalities through taxes or whatever. However, don't think that government oversight is a panacea - corruption and pressure to deliver are everywhere.
The chromium source is already enormous - upwards of 500MB. Building it takes forever. Why? Simple - Google bundles every library already on your system in the source.
Now it looks like they're tossing flash on the pile.
Maybe we should throw in OpenOffice so that we can edit documents simply by browsing to them. Plus, then if I have openoffice open outside of chromium they can each use half a gig of RAM since the OS can't share the pages!
What is wrong with using shared libraries and plugins? If Google wants to write their own flash plugin and maybe bundle it by default, I have no issues with that. However, there is no reason this needs to be hard-coded into the browser.
I was thinking about that also - only thing I could see is being able to maneuver in-between dropping kill vehicles, but that would only save a little per-vehicle weight. The only reason to recover the launcher would be if it were unusually complex, but I can't see that either.
Satellite capture might make sense, but then why have wings? If anything landing with all that extra weight would only be harder with wings.
To capture a satellite would require being in the same inclination as the target, so we're now at low inclination. At low inclination I'm not sure the wings buy you a whole lot.
To be honest, wings don't buy a whole lot, period. So, I'm hard-pressed to figure out what they're for. Maybe if you had a scram jet you could use them for inclination changes - just aerobrake and then re-insert. That is only effective if the scrams buy you a lot of cheap energy.
I'd actually think that something like this would be ideal to TAKE OUT a satellite, or a satellite interceptor.
A polar orbit means that its relative velocity would be large compared to any less inclined orbit.
Ability to launch and return in a single polar orbit means that it would be hard to shoot down - it would have to fly right over an enemy launch site to do so since they wouldn't spot it until it was entering their airspace and there would be no time to vector an interception from elsewhere. You only have a few minutes to launch even if you happen to have an ASAT missile right on its flight path (which obviously the US would avoid anyway when they put it into orbit).
So, the USAF identifies a bunch of satellites they want to shoot down, then they put this thing into orbit which parks interceptors in polar orbits that will hit each of the targets. Then it re-enters and returns to base.
Another option is recon - this thing could be launched at any inclination to get to any point in the earth quickly and then be able to return to base more quickly with cross-range capability.
Those are just some wild guesses. Wings do give you options - no sense having them unless your mission demands them.
In short, there are very good reasons for the rules that you may think are stupid, oppressive, etc.
I don't think he ever claimed that there wasn't a reason for them, although I'd debate whether the rules aren't still "stupid."
This guy's story is exactly the reason why information shouldn't be withheld from jurors to the degree that it is. He knew about the other charges, and yet he still weighed the evidence and found it insufficient to sustain a conviction.
Maybe some jurors would be too dumb to be the same. I'm not quite sure why we let them sit on panels if we think they're so dumb in the first place.
If I were on a jury I'd feel bound to do what is right. That certainly would include not finding somebody guilty of a crime as defined by the judge if there were not evidence to sustain this charge, but it would not necessarily be limited to only this responsibility. In any case, as a working person with half a brain I doubt I'll ever serve on a jury again.
The other problem in california was that they invented a really stupid market system.
Suppose a distributor had agreements to deliver 1000MWh of electricity for 5 cents/kWh. Demand goes up and they need to buy 1 more MWh. The only guy willing to sell it wants $100/kWh.
In a normal market they'd have a few choices: 1. Just institute rolling blackouts and not pay for the extra power. 2. Pay $100k to buy 1 MWh at that price.
In the CA market, the regulators forced the distributors to buy the power, so #1 wasn't an option. They also forced the distributor to pay everybody the same, so not only did they fork out $100k for the the extra 1MWh, but they also had to pay everybody else who had already agreed to 5 cents/kWh an extra $99.95/kWh, or $100M in total.
So, that one supplier effectively increased the costs of the distributor by 2000x with a single bid on 0.1% of the supply.
That is one messed up "deregulation" scheme...
All that said, power is one of many areas where a healthy amount of regulation makes sense, just due to its nature. The problem in CA wasn't regulation, or deregulation. The problem was that a bunch of rich people bribed a bunch of politicians to put their interests ahead of the population. That happens all the time among conservatives, liberals, free-market types, and command-economy types. Corruption is a very universal thing.
The problem is that if you try to advertise a route for a/24 all those big routers will just ignore you.
Selling IPv4 addresses only works if you sell them to somebody on your subnet, or if you sell a LOT of these addresses. Your other option is to offer to tunnel traffic.
You can't just take 10 IPs you aren't using in California and ebay them off to some guy in Denmark unless you have some way of actually directing traffic to him.
Agreed - I use vmware-server, and it is annoying that their remote console viewer plugins don't work with chrome.
So, the firefox people will just say "install firefox!" So, that's wonderful until you discover that their plugins don't work on the most recent release of firefox either. You need to install some old v2-based firefox browser to get it to work.
A standards-compliant plugin that works in all browsers also will tend to work across browser versions. That's why I don't want to see majority market share in this area. If there are 5 major browsers then everybody needs to figure out how to make their stuff work with all of them, and that means standards.
I feel the same way at work - I don't like Oracle-only applications even though we run Oracle. Why? Simple, an Oracle-only application is likely to choke every time Oracle changes any detail in their implementation. An ANSI SQL-based app already needs to be flexible, and that means that you're less likely to have it break every time a vendor releases an upgrade because those basic standards don't change much.
The odd thing is that file transfers like used with torrents should be fairly friendly to constant-bandwidth approaches like what mixmaster used. It probably wouldn't be any slower than tor is today. If anything, having people downloading stuff might help since it gives the network more traffic to mix stuff in.
What is needed, however, is to build the anonymity into the file transfer protocol, so that you don't need exit nodes. The reason torrent over tor is popular is that you can use existing swarms with it (with thousands of seeds sometimes) - if you try to look on something like freenet or i2c for a file you'd be lucky to find the US declaration of independence, with maybe one "seed."
The big problem is the network effect - any new sharing network is going to have fewer contributers than anything established, which makes it hard to change the status quo.
The tor folks should probably go out of their way to create/promote such a new network. It could be used to handle non-web video traffic, and it would get all that bandwidth off of tor!
The "business-oriented" hardware often is analogous to LTS OS software. Big box manufacturers usually change parts any time they can get a deal, so their systems can vary considerably within a given model. Their business-oriented series do not do this - they'll stick with one set of components for a very long time, and guarantee to make compatible parts available for a long time as well. And I mean like-for-like parts, not semi-equivalent parts.
That means that when your video card dies in a business-series PC you can just reach onto your spare shelf, pop in a new one, and boot the PC up. No need to update drivers or reconfigure, since the old card is identical to the new one.
Sure, you are paying new-card prices for old cards, but in many of these industries a simple supply chain is much more useful than cutting-edge performance.
Business-oriented suppliers also will have offerings like $1 leases, and leases/warranties with a provision that allows hard drives to not be returned. They'll usually also be willing to pre-image it with any image you supply.
The bottom line with most of these services is that you can put a once-and-done pricetag on a PC purchase, or get close to that. This simply goes on the depreciation schedules and the accountants are happy.
Rather, just make chrome use an extensible plugin api for video playback. Why is it that support for codecs needs to be hard-coded into a browser anyway? If I have a codec for bob's-crazy-video-standard, the browser should just work...
Here is an interesting question: How many missiles should one fire to maximize your chance of a hit?
With one missile you maximize your chances of the missile getting pretty close to the target before they even realize they're under attack.
With a barrage you increase the chance that your attack will be spotted further out and therefore subject your barrage to a larger counter-defense, but you increase the probability of at least one missile making it through.
More isn't necessarily better. Before an alarm is sounded ships might have their air search radars off for stealth, or might have weapons masked (many weapons systems do not have a 360 degree field of fire due to obstructions - although I suspect that vertical launch systems may not have this flaw). If somebody spots a missile, then every ship will go to full speed, angle for maximum defensive posture, fire countermeasures, etc. A single missile just might sneak in close enough that nobody can react in time.
In any case, I do agree that it only makes sense to develop countermeasures against the kinds of scenarios you envision. It would be foolish to simply ignore these threats.
Yup, although I could certainly see linux running in the firmware of some digital audio processor or mixing console. Most likely that would bypass any of the actual sound platforms in linux and would just manage bits and bytes. From what I've seen the actual sound platforms on linux are a huge mess.
The navy would be an obvious source if ideas here. Ships have had watches for centuries, and a watch that doesn't see anything isn't any better than no watch at all. I suspect that modern navies rotate them frequently as a result.
Yeah, but who wants to take one shot at a US carrier? If you're going to start a war you better make sure you can finish it! Most 3rd world dictators don't share advanced weapons systems with terrorists for the same reason. I think that until these become as cheap and ubiquitous as RPGs that they'll be outside the reach of terrorists.
A guy on a fishing boat who can chart his location within a mile with a satellite radio would be sufficient intel to launch a cruise missile attack. Of course the missile is still subject to countermeasures, just like a saturation missile attack from a bunch of Tu-22s in 1989 (which is the whole reason Aegis was invented).
After the war starts, the exclusion zone would be sufficient to keep fishing boats away most likely, and even the launch platform (these missiles only go a few hundred miles, so just keep container ships that far away and you're fine - or at least you have a lot more warning).
I'd forget about submersibles. The only reason fishing boats could get close would be because nobody wants to mess with them - not because they're stealthy. A submersible is automatically suspicious and there is no way anything not invented by a fairly major power is going to sneak in close to a carrier. US ASW defenses are intended to defeat very quiet diesel boats made by the old USSR. In a war situation anything underwater would be treated as a warship so if it isn't supposed to be there it won't stay there for long.
Ok, so you check your position every 30 minutes. Anything else?
I've done the whole flight sim across the atlantic bit too - it is REALLY boring if you don't use time acceleration. Real world procedures are a little better since I think you're supposed to keep track of your position relative to your divert fields in case you lose an engine, but you're still talking about something to do every 10-20 minutes at most. You're more likely to do that if you just set your watch to beep every ten minutes and then find something to occupy yourself during the remainder.
Yup - there was an article posted here ages ago about an X-ray system for airports that randomly would add in images of guns/knives/etc for the operator to find. The operator would hit a button when they spotted the item and it would disappear from the screen. Of course, if it were a real weapon it wouldn't disappear and then the operator would sound the alarm.
The machine then kept score. A low score indicated a distracted operator and signaled a supervisor.
Instead of staring blankly at a screen that contained nothing 99.999% of the time the operator would be engaged in actually finding things maybe 10% of the time.
A plane could do the same thing - it could from time to time cause an indicator to go out of range or something like that and the pilot would have to spot it. Or it could have sanctioned games built in or something like you suggest.
Humans can't just stare at blue sky and unmoving gauges for hours on end - we're not wired for it. Simply threatening to fire people doesn't work. You need to give their brains something to do.
Actually, you pretty-much HAVE to have a tracking sensor on the missile itself if you want to hit anything.
However, if you fire off a missile in a random direction you're going to blow up some ship at random at best - most likely a tanker. Also, radar is subject to jamming, and you bet that if somebody spots a radar emitter travelling at mach 3 they're going to jam it! IR or visual is only useful if you have a very good idea of where the target is - it can't search a large area. Even radar on a missile will only spot something in a cone maybe a few miles across, which isn't much even in the Persian Gulf.
Keep in mind that a cruise missile is exactly the kind of threat that US naval defenses were designed to counter. The only twist here is sneaking one onto a merchant ship.
This sounds like the Win NT remediation at work (guess what, we still have that too). Corporate IT comes in and says "thou shalt get rid of NT," business points to a huge piece of automated equipment whose software only runs on NT. Corporate IT says "upgrade the software" - business says no upgrade is available. Corporate IT does some research and says "gee, the vendor says they support XP if you just buy automated equipment Mark III" - business points out that automated equipment Mark I cost $500k per unit, and we have 200 units deployed, and the business justification was based on a 15-year depreciation schedule, three years ago. CIO gets phone call from VP in the business, and Corporate IT was told to figure out a practical way to mitigate the risk without getting rid of the controllers.
IE6 is just the same thing over again. Deploying software in a huge company involves all kinds of costs, and upgrading that software has HUGE transition costs. If that old data has a retention period (virtually always the case) you also need to make sure it remains accessible. Most likely the new software and the old software don't do exactly the same thing, so you have all kinds of additional gaps to address.
IT in big companies is messy - long-term support doesn't mean three years, and even ten can be painful.
All of these things require getting somebody within visual or radar range of the target. I doubt that small boats are allowed within visual or radar range of carriers.
At best you'd get one shot, and then a war with the US. If the US were actually in a shooting war it would simply declare a huge exclusion zone and sink anything inside of it that wasn't under US escort (merchant shipping would be friendly-flag only, with full military coordination).
The countries that could even hold out for a little against a US onslaught already have superior weapons systems, and this won't shift the balance of power for those countries that can't.
The carrier will still travel more than one length in the time the missile takes to fly, and GPS guidance only takes a missile to a geographic location. If you use GPS guidance ONLY then the missile can only travel to the predicted location of the ship. If the ship is even 10 feet away from that precise location in space, then the missile won't cause significant damage to it other than killing a few people standing on the deck at most.
To actually hit the ship the missile needs to actually "see" the ship - whether IR or radar or whatever. GPS guidance only can get the missile to the vicinity of the ship.
In the real world the problem is even more difficult - the ship's exact location may not be known, and it will likely move between sighting and firing, unless it is sighted directly by the launching platform. Plus, the original location of the target will be only accurate to within maybe 100 yards or so. That is plenty of good to get a missile close enough to find a ship with radar, but not good enough to hit a moving target with a missile with no terminal guidance.
Now, land-based targets are an entirely different matter. A building doesn't move, so GPS alone is fine to hit it. Plus, land-based maps usually are more accurate. A ship in the middle of an ocean has no navigational reference points nearby, so its location assignment is limited by the accuracy of the servos on the sensor that located it. A tank parked in a parking lot can be measured relative to a street intersection on the same photo whose exact location is known.
Slashdot Mirror
Yup - clinical trials cost a fortune. Most of the costs of developing a drug are in the trials - tens of millions of dollars each and hundreds of millions of dollars in total (and that is when they're run assembly-line style by organizations that do nothing but run and supply trials 24x7).
I think half of it is essentially corporate welfare for doctors and hospitals. The costs you point out are very real of course. You also need to factor in that if you pay a doctor $100 per visit per subject and somebody else offers $150 per visit per subject then you won't get any subjects. There seems to be a lot more demand for subjects than supply. Of course, the actual volunteers don't see a dime of this money, and chances are their doctor doesn't mention to them that they are recommending experimental treatment A over experimental treatment B because they get more money out of it. Not that doctors would actually do anything like base treatment decisions on their financial interests...
Actually, if you look at the FDA's debarment list (a list of people forbidden to work on pharmaceuticals or trials), you'll find most are doctors. The most typical offense is signing up people for trials who shouldn't be in the trials so that they can get paid all the associated fees. This is bad for everybody. The patient gets experimental drugs that may not help them, or might even hurt them. The pharmaceutical company gets clinical trial data that contains more noise, which means they have to subject even more people to experimental medication to find out if it works or not, and pay more money as a result. The public ends up with less accurate clinical trial data, which might mean that a drug that actually works doesn't get approved (which means people don't benefit from it, or maybe they keep taking inferior drugs that are actually less safe even though the trial data suggests otherwise).
Human clinical trials are a mess. Their ability to generate meaningful data is also very weak compared to things like animal testing. The problem is that they're all we have, so we need to make the most of them.
The GP is right that the only way we can see independent trials happen is if the government funds them - or maybe VERY LARGE charities. I'm all for it - if an independent trial might benefit the public by showing benefit from a cheap "new" drug, or nail the coffin in a bad drug, then let's have it. And I'm generally in favor of drug patents. Competition is always good for consumers and it keeps private companies more honest.
Agreed.
Also - if cars drove themselves we wouldn't need so many of them. My car could drop me off at work, and then drive home to take care of my wife all day until I need it again.
If I needed a car unexpectedly I'd just punch something into my smartphone and in 2 mins a rental/taxi would drive up to pick me up.
Public transit would make a whole lot more sense if a car dropped me off on the train platform, and another was sitting there ready to pick me up for the last mile when I got off at the other end.
For a big family vacation I could have a second car just shuttle my luggage. Maybe it would drive at 45mph for maximum economy and leave early or arrive late.
No need to own a big truck, but I could have one with two minutes notice if I need to haul something. If I need to drop something off at a neighbor's there would be no need for me to even ride along for the trip as long as they're to unload at the destination.
Stores wouldn't need parking lots. Each town would have big parking garages strategically located, and cars would drop people off and go find something to do with themselves until needed. Just think about how much less asphault we'd need, and how much that would be better for the enviornment/etc.
Oh, no cars idling anywhere either, no red lights, etc. That has to be great for improving gas mileage.
Lots of options if you get rid of the human drivers. For far less than we spend on all the problems caused by our current transportation system we could probably fix it...
I'm not sure we're just talking about education - we're talking about reaction times and fallibility.
Why shouldn't my car travel an average speed of less than 45mph on minor roads without stops on my way to work? Right now I probably average 20mph tops - 45mph when driving, and 0mph at lights/etc.
Automated vehicles would just proceed at fairly high speeds the entire distance of a trip with greatly reduced risk of accidents/etc. On highways they would cruise at 100mph or so, with a few inches clearance between cars to maximize aerodynamics. Presumably cars following the same routes would huddle together, ordering themselves to minimize aerodyanmic profile.
You can't mix this kind of system with human drivers, regardless of how much training they have. What will you do when you come to your first intersection between two four-lane roads without a traffic light with traffic moving through at highway speeds in all directions simultaneously? With automated cars each car would get a reservation for every spot along the intersection, with sufficient margins for error/failures/etc. No human could safely navigate under these conditions.
The real issues (as others have pointed out) are the liabilities. Today Ford's cars kill maybe 1k people per year, but they have no liability. If they automated their cards and killed only two people per year they'd be sued for everything they own. Our legal system doesn't care that they've saved 998 people per year - only that they are now subject to some kind of liability for the 2 they didn't manage to save.
Agreed.
Utilities should be BORING industries. By all means they should improve bandwidth/etc, and be able to sell that. However, they should be pipes and wires.
The problem is all the vertical integration. Nobody wants to be a company that guarantees shareholders 25cents per share every quarter from now until eternity. They want to be able to promise double-digit earnings growth, and that requires the ability to grow markets. However, when your market is ever home in a 10 mile radius who could own a phone, and they already own phones, then there is no room to grow.
The problem is greed. It isn't like the CEO of a boring utility company doesn't make a good six-figure salary. However, who wants 6-figures when they could be the next Bill Gates? Well, if you want to leave and start your own company that's fine, but when you want to use a government-granted monopoly as the springboard for world conquest then don't be surprised when taxpayers start complaining.
Yup - we should get the government to handle this because they have such a better track record in managing unusual risks, like they did with the toxic asset crash and the space shuttle.
Don't get me wrong - we do need more oversight, and we do need to make companies cover their externalities through taxes or whatever. However, don't think that government oversight is a panacea - corruption and pressure to deliver are everywhere.
The chromium source is already enormous - upwards of 500MB. Building it takes forever. Why? Simple - Google bundles every library already on your system in the source.
Now it looks like they're tossing flash on the pile.
Maybe we should throw in OpenOffice so that we can edit documents simply by browsing to them. Plus, then if I have openoffice open outside of chromium they can each use half a gig of RAM since the OS can't share the pages!
What is wrong with using shared libraries and plugins? If Google wants to write their own flash plugin and maybe bundle it by default, I have no issues with that. However, there is no reason this needs to be hard-coded into the browser.
I was thinking about that also - only thing I could see is being able to maneuver in-between dropping kill vehicles, but that would only save a little per-vehicle weight. The only reason to recover the launcher would be if it were unusually complex, but I can't see that either.
Satellite capture might make sense, but then why have wings? If anything landing with all that extra weight would only be harder with wings.
To capture a satellite would require being in the same inclination as the target, so we're now at low inclination. At low inclination I'm not sure the wings buy you a whole lot.
To be honest, wings don't buy a whole lot, period. So, I'm hard-pressed to figure out what they're for. Maybe if you had a scram jet you could use them for inclination changes - just aerobrake and then re-insert. That is only effective if the scrams buy you a lot of cheap energy.
I'd actually think that something like this would be ideal to TAKE OUT a satellite, or a satellite interceptor.
A polar orbit means that its relative velocity would be large compared to any less inclined orbit.
Ability to launch and return in a single polar orbit means that it would be hard to shoot down - it would have to fly right over an enemy launch site to do so since they wouldn't spot it until it was entering their airspace and there would be no time to vector an interception from elsewhere. You only have a few minutes to launch even if you happen to have an ASAT missile right on its flight path (which obviously the US would avoid anyway when they put it into orbit).
So, the USAF identifies a bunch of satellites they want to shoot down, then they put this thing into orbit which parks interceptors in polar orbits that will hit each of the targets. Then it re-enters and returns to base.
Another option is recon - this thing could be launched at any inclination to get to any point in the earth quickly and then be able to return to base more quickly with cross-range capability.
Those are just some wild guesses. Wings do give you options - no sense having them unless your mission demands them.
In short, there are very good reasons for the rules that you may think are stupid, oppressive, etc.
I don't think he ever claimed that there wasn't a reason for them, although I'd debate whether the rules aren't still "stupid."
This guy's story is exactly the reason why information shouldn't be withheld from jurors to the degree that it is. He knew about the other charges, and yet he still weighed the evidence and found it insufficient to sustain a conviction.
Maybe some jurors would be too dumb to be the same. I'm not quite sure why we let them sit on panels if we think they're so dumb in the first place.
If I were on a jury I'd feel bound to do what is right. That certainly would include not finding somebody guilty of a crime as defined by the judge if there were not evidence to sustain this charge, but it would not necessarily be limited to only this responsibility. In any case, as a working person with half a brain I doubt I'll ever serve on a jury again.
The other problem in california was that they invented a really stupid market system.
Suppose a distributor had agreements to deliver 1000MWh of electricity for 5 cents/kWh. Demand goes up and they need to buy 1 more MWh. The only guy willing to sell it wants $100/kWh.
In a normal market they'd have a few choices:
1. Just institute rolling blackouts and not pay for the extra power.
2. Pay $100k to buy 1 MWh at that price.
In the CA market, the regulators forced the distributors to buy the power, so #1 wasn't an option. They also forced the distributor to pay everybody the same, so not only did they fork out $100k for the the extra 1MWh, but they also had to pay everybody else who had already agreed to 5 cents/kWh an extra $99.95/kWh, or $100M in total.
So, that one supplier effectively increased the costs of the distributor by 2000x with a single bid on 0.1% of the supply.
That is one messed up "deregulation" scheme...
All that said, power is one of many areas where a healthy amount of regulation makes sense, just due to its nature. The problem in CA wasn't regulation, or deregulation. The problem was that a bunch of rich people bribed a bunch of politicians to put their interests ahead of the population. That happens all the time among conservatives, liberals, free-market types, and command-economy types. Corruption is a very universal thing.
The problem is that if you try to advertise a route for a /24 all those big routers will just ignore you.
Selling IPv4 addresses only works if you sell them to somebody on your subnet, or if you sell a LOT of these addresses. Your other option is to offer to tunnel traffic.
You can't just take 10 IPs you aren't using in California and ebay them off to some guy in Denmark unless you have some way of actually directing traffic to him.
Agreed - I use vmware-server, and it is annoying that their remote console viewer plugins don't work with chrome.
So, the firefox people will just say "install firefox!" So, that's wonderful until you discover that their plugins don't work on the most recent release of firefox either. You need to install some old v2-based firefox browser to get it to work.
A standards-compliant plugin that works in all browsers also will tend to work across browser versions. That's why I don't want to see majority market share in this area. If there are 5 major browsers then everybody needs to figure out how to make their stuff work with all of them, and that means standards.
I feel the same way at work - I don't like Oracle-only applications even though we run Oracle. Why? Simple, an Oracle-only application is likely to choke every time Oracle changes any detail in their implementation. An ANSI SQL-based app already needs to be flexible, and that means that you're less likely to have it break every time a vendor releases an upgrade because those basic standards don't change much.
The odd thing is that file transfers like used with torrents should be fairly friendly to constant-bandwidth approaches like what mixmaster used. It probably wouldn't be any slower than tor is today. If anything, having people downloading stuff might help since it gives the network more traffic to mix stuff in.
What is needed, however, is to build the anonymity into the file transfer protocol, so that you don't need exit nodes. The reason torrent over tor is popular is that you can use existing swarms with it (with thousands of seeds sometimes) - if you try to look on something like freenet or i2c for a file you'd be lucky to find the US declaration of independence, with maybe one "seed."
The big problem is the network effect - any new sharing network is going to have fewer contributers than anything established, which makes it hard to change the status quo.
The tor folks should probably go out of their way to create/promote such a new network. It could be used to handle non-web video traffic, and it would get all that bandwidth off of tor!
The "business-oriented" hardware often is analogous to LTS OS software. Big box manufacturers usually change parts any time they can get a deal, so their systems can vary considerably within a given model. Their business-oriented series do not do this - they'll stick with one set of components for a very long time, and guarantee to make compatible parts available for a long time as well. And I mean like-for-like parts, not semi-equivalent parts.
That means that when your video card dies in a business-series PC you can just reach onto your spare shelf, pop in a new one, and boot the PC up. No need to update drivers or reconfigure, since the old card is identical to the new one.
Sure, you are paying new-card prices for old cards, but in many of these industries a simple supply chain is much more useful than cutting-edge performance.
Business-oriented suppliers also will have offerings like $1 leases, and leases/warranties with a provision that allows hard drives to not be returned. They'll usually also be willing to pre-image it with any image you supply.
The bottom line with most of these services is that you can put a once-and-done pricetag on a PC purchase, or get close to that. This simply goes on the depreciation schedules and the accountants are happy.
NO - don't push it into chrome!
Rather, just make chrome use an extensible plugin api for video playback. Why is it that support for codecs needs to be hard-coded into a browser anyway? If I have a codec for bob's-crazy-video-standard, the browser should just work...
Sure.
Here is an interesting question: How many missiles should one fire to maximize your chance of a hit?
With one missile you maximize your chances of the missile getting pretty close to the target before they even realize they're under attack.
With a barrage you increase the chance that your attack will be spotted further out and therefore subject your barrage to a larger counter-defense, but you increase the probability of at least one missile making it through.
More isn't necessarily better. Before an alarm is sounded ships might have their air search radars off for stealth, or might have weapons masked (many weapons systems do not have a 360 degree field of fire due to obstructions - although I suspect that vertical launch systems may not have this flaw). If somebody spots a missile, then every ship will go to full speed, angle for maximum defensive posture, fire countermeasures, etc. A single missile just might sneak in close enough that nobody can react in time.
In any case, I do agree that it only makes sense to develop countermeasures against the kinds of scenarios you envision. It would be foolish to simply ignore these threats.
Yup, although I could certainly see linux running in the firmware of some digital audio processor or mixing console. Most likely that would bypass any of the actual sound platforms in linux and would just manage bits and bytes. From what I've seen the actual sound platforms on linux are a huge mess.
The navy would be an obvious source if ideas here. Ships have had watches for centuries, and a watch that doesn't see anything isn't any better than no watch at all. I suspect that modern navies rotate them frequently as a result.
Yeah, but who wants to take one shot at a US carrier? If you're going to start a war you better make sure you can finish it! Most 3rd world dictators don't share advanced weapons systems with terrorists for the same reason. I think that until these become as cheap and ubiquitous as RPGs that they'll be outside the reach of terrorists.
A guy on a fishing boat who can chart his location within a mile with a satellite radio would be sufficient intel to launch a cruise missile attack. Of course the missile is still subject to countermeasures, just like a saturation missile attack from a bunch of Tu-22s in 1989 (which is the whole reason Aegis was invented).
After the war starts, the exclusion zone would be sufficient to keep fishing boats away most likely, and even the launch platform (these missiles only go a few hundred miles, so just keep container ships that far away and you're fine - or at least you have a lot more warning).
I'd forget about submersibles. The only reason fishing boats could get close would be because nobody wants to mess with them - not because they're stealthy. A submersible is automatically suspicious and there is no way anything not invented by a fairly major power is going to sneak in close to a carrier. US ASW defenses are intended to defeat very quiet diesel boats made by the old USSR. In a war situation anything underwater would be treated as a warship so if it isn't supposed to be there it won't stay there for long.
Ok, so you check your position every 30 minutes. Anything else?
I've done the whole flight sim across the atlantic bit too - it is REALLY boring if you don't use time acceleration. Real world procedures are a little better since I think you're supposed to keep track of your position relative to your divert fields in case you lose an engine, but you're still talking about something to do every 10-20 minutes at most. You're more likely to do that if you just set your watch to beep every ten minutes and then find something to occupy yourself during the remainder.
Yup - there was an article posted here ages ago about an X-ray system for airports that randomly would add in images of guns/knives/etc for the operator to find. The operator would hit a button when they spotted the item and it would disappear from the screen. Of course, if it were a real weapon it wouldn't disappear and then the operator would sound the alarm.
The machine then kept score. A low score indicated a distracted operator and signaled a supervisor.
Instead of staring blankly at a screen that contained nothing 99.999% of the time the operator would be engaged in actually finding things maybe 10% of the time.
A plane could do the same thing - it could from time to time cause an indicator to go out of range or something like that and the pilot would have to spot it. Or it could have sanctioned games built in or something like you suggest.
Humans can't just stare at blue sky and unmoving gauges for hours on end - we're not wired for it. Simply threatening to fire people doesn't work. You need to give their brains something to do.
Actually, you pretty-much HAVE to have a tracking sensor on the missile itself if you want to hit anything.
However, if you fire off a missile in a random direction you're going to blow up some ship at random at best - most likely a tanker. Also, radar is subject to jamming, and you bet that if somebody spots a radar emitter travelling at mach 3 they're going to jam it! IR or visual is only useful if you have a very good idea of where the target is - it can't search a large area. Even radar on a missile will only spot something in a cone maybe a few miles across, which isn't much even in the Persian Gulf.
Keep in mind that a cruise missile is exactly the kind of threat that US naval defenses were designed to counter. The only twist here is sneaking one onto a merchant ship.
This sounds like the Win NT remediation at work (guess what, we still have that too). Corporate IT comes in and says "thou shalt get rid of NT," business points to a huge piece of automated equipment whose software only runs on NT. Corporate IT says "upgrade the software" - business says no upgrade is available. Corporate IT does some research and says "gee, the vendor says they support XP if you just buy automated equipment Mark III" - business points out that automated equipment Mark I cost $500k per unit, and we have 200 units deployed, and the business justification was based on a 15-year depreciation schedule, three years ago. CIO gets phone call from VP in the business, and Corporate IT was told to figure out a practical way to mitigate the risk without getting rid of the controllers.
IE6 is just the same thing over again. Deploying software in a huge company involves all kinds of costs, and upgrading that software has HUGE transition costs. If that old data has a retention period (virtually always the case) you also need to make sure it remains accessible. Most likely the new software and the old software don't do exactly the same thing, so you have all kinds of additional gaps to address.
IT in big companies is messy - long-term support doesn't mean three years, and even ten can be painful.
All of these things require getting somebody within visual or radar range of the target. I doubt that small boats are allowed within visual or radar range of carriers.
At best you'd get one shot, and then a war with the US. If the US were actually in a shooting war it would simply declare a huge exclusion zone and sink anything inside of it that wasn't under US escort (merchant shipping would be friendly-flag only, with full military coordination).
The countries that could even hold out for a little against a US onslaught already have superior weapons systems, and this won't shift the balance of power for those countries that can't.
The carrier will still travel more than one length in the time the missile takes to fly, and GPS guidance only takes a missile to a geographic location. If you use GPS guidance ONLY then the missile can only travel to the predicted location of the ship. If the ship is even 10 feet away from that precise location in space, then the missile won't cause significant damage to it other than killing a few people standing on the deck at most.
To actually hit the ship the missile needs to actually "see" the ship - whether IR or radar or whatever. GPS guidance only can get the missile to the vicinity of the ship.
In the real world the problem is even more difficult - the ship's exact location may not be known, and it will likely move between sighting and firing, unless it is sighted directly by the launching platform. Plus, the original location of the target will be only accurate to within maybe 100 yards or so. That is plenty of good to get a missile close enough to find a ship with radar, but not good enough to hit a moving target with a missile with no terminal guidance.
Now, land-based targets are an entirely different matter. A building doesn't move, so GPS alone is fine to hit it. Plus, land-based maps usually are more accurate. A ship in the middle of an ocean has no navigational reference points nearby, so its location assignment is limited by the accuracy of the servos on the sensor that located it. A tank parked in a parking lot can be measured relative to a street intersection on the same photo whose exact location is known.