When you think about it, this is the only sensible approach. Do you want every municipality that owns a helicopter to be trying to police aircraft that are flying overhead? Maybe the plane's registration is bogus. Great, call it in to the FAA as a good citizen and let them deal with it.

There are a lot of safety issues when you try to deal with issues in the air. Indeed, I've heard ATC recordings where ATC is basically trying to ream somebody out for not following procedures correctly, and that is also something that shouldn't happen. Deal with problems on the ground - if somebody violates the rules they should record it and refer it for enforcement action and write up a report. Trying to deal with problems in the air just means you're putting others at risk by not doing your job.

I'm all for enforcement, but at the right time and place.

Agree. The US even has the FCPA - it is outright illegal to bribe foreign officials. That law isn't enforced as well as it probably could be, but it is enforced and you do hear about a scandal from time to time. I know that my employer trains on the act and makes compliance a clear policy (though I have no idea how much they follow-through in practice - I wouldn't be privy to enforcement actions).

I'm not sure to what degree this is the case in other countries.

Yeah, I was thinking something much smaller/lighter. Almost something fieldable by an amateur (the telescope itself would be in the amateur range, but the drone would be a bit large if you wanted it to go high-altitude.

I've read in astronomy forums (a while ago) that there is a lot of emphasis on really big telescope projects, but for much less money you can get a lot of science done if you take a much smaller instrument but manage it professionally (automation, calibration, etc). It isn't unlike in computing where there is a place for both the big fancy supercomputer with high memory bandwidth and super-fast CPU, and the massively parallel commodity cluster. Different problems are best solved with either approach. Maybe the Hubble can image galaxies at the edge of the visible universe, but for the same price you could have 20,000 cheap telescopes doing continuous surveys of large regions of the sky. Each has its place.

For the 3rd world, though, the A10 seems risky. Its main advantage is getting in low and slow and basically fighting "melee" against batallions of vehicles.

For what it does it is a GREAT platform, but you wouldn't send a lot of these against an enemy and not expect losses. They've vulnerable to fire from MANPADs and such which even terrorist groups could have.

What they can do is fly in at 300 feel where the nearby SA-10 site won't see them and drop bombs, while the F15s would be shot down 80 miles away by the SAM.

If the enemy doesn't have advanced SAMs, then the F15 can just loiter at 40k feet and drop a bomb when requested. Nothing a terrorist has is going to be able to touch it, and if there are older SAMs those can be effectively destroyed at the start of the mission.

I agree that drones are really where everybody should be going. Just look at them like multi-stage cruise missiles. A drone would have a lower radar cross section due to its size, it WOULD be cheaper/faster to mass-produce/replace, and it could fly the terminal portion of its mission at very low altitude. It only costs money, so fighting a war of attrition is politically acceptable - losing 1000 drones in a fight isn't the same as losing 1000 pilots with the POWs being fought over for decades later. It also enables distraction tactics that involve sacrificing aircraft.

SSL goes beyond the naivety of government trust. It also suffers from what amounts to a global namespace/trust/etc issue.

Any CA can issue a certificate for any domain, a domain generally can only have one certificate, and the trusted CA list is managed by the browser, not the user.

So, if you trust your government (naievely), and distrust everybody else, it won't work. Your browser will constantly be wanting to add CAs you don't trust, and might not include ones you trust. Then, if you drop a bunch of CAs then a bunch of websites won't work. A website doesn't have the option of getting certificates from 14 different CAs so as to be trusted by everybody - they have to pick one and everybody has to trust them.

So, users are basically forced to accept CAs they've never heard of, and the whole system is a mess as a result.

That makes me wonder if you couldn't stick a somewhat decent telescope on top of a drone. I'm talking something in the 12-18" range - not some monster that should be on top of a mountain. There is probably a lot of science that could be done with a very good quality small mirror with little atmosphere and always above the clouds. If you wanted to go really high you'd need turbines though - so it would be reasonably large (need lots of fuel, relatively speaking). Stability would be important - one of the advantages of no atmosphere is that you can take very long exposures without much background.

It depends on the objectives. If your goal is to annex them, no.

If your goal is to deter a conflict, or fight a conflict over some remote piece of land where the war could possibly be contained, then maybe. Nobody has incentive to start a nuclear war, so it would likely stay conventional. Meaningful missile defense might also have an impact on the calculus - if a small nuclear strike could be reliably defeated that would probably make a foe less likely to try to escalate to a regional nuclear conflict (all in is just suicide).

However, it would be really costly for all sides, so the interest would have to be something actually worth fighting for. The US could probably deter a Russian invasion of Ukraine, for example, by stationing forces in Lithuania sufficient to threaten Moscow. The relative strength of both sides would probably allow a US victory (crazy thought), if the Russians didn't kill everybody on Earth, but it would depend on them actually allowing a US buildup, since we're separated by an ocean and resupply in that area would be very threatened by Russian Navy/Air/etc. Most likely there would be a Naval war, which could result in attacks on US ports/interior/etc, but on the whole the Russians don't have much of a navy these days and they wouldn't last too long.

But, like I said it would have to be an interest worth fighting for and the US really isn't going to risk even conventional bombs going off in DC and Manhattan and Saudi Arabia over Ukraine. So, all the above is really just a fantasy scenario. Maybe if somebody like Russia went really nuts going after US vital interests it would come to war, but that seems rather unlikely. But, on the other hand having modern weapons platforms only makes it less likely.

He has some valid points (I only listed to most of the first part), but I'm not sure it is entirely current. For example, he claims that it is trivial for maneuverable aircraft to dodge SAMs. That was true back in the Vietnam days (though we still lost many), and true in Iraq (which was basically Modern US vs Vietnam-era Iraq), but if we go into a modern war those fighters will have to dodge the likes of SA-10s. The missile sensors are much harder to fool, and they have thrust vectoring and are very light at their terminal stage which basically makes them potentially very hard to evade.

Even stealth aircraft are unproven in that scenario since nobody reveals their full capabilities. We'll never know until we actually get into a war with somebody who can actually hurt us, instead of bombing countries that are 30 years behind.

The US likes to maintain foreign interests like Eastern Europe, Taiwan, etc that are very far away from us but fairly close to whoever they're likely to get into conflict with. That means that the US will need all the capability it can get in a war. Since our fighters will be tanker-supported from fields hundreds of miles away, we'll probably be outnumbered, and fighting near foreign territory. Also, the US likes to meddle and voters don't like to hear about dead pilots.

The A10 was a wonder of an aircraft, but I'm not entirely sure it would still be useful in a serious war.

Against ISIS it might be useful, assuming that they're fighting in open terrain. An A10 is not particularly useful against targets in the middle of a city - if anything an aircraft that can loiter at altitude might be more useful for dropping bombs on designated buildings - nobody is going to be swooping over city blocks and hitting the right building visually.

Really the advantage of the A10 over other aircraft is that it can potentially use terrain masking to protect itself from large SAM emplacements. That doesn't matter in a place like Iraq. It could matter in a serious war with a modern defender. The reality is that we don't know how effective modern SAMs will be against things like the F22, and we also don't know how effective aircraft like the A10 will be at getting past them. The former is going to depend greatly on the F22's stealthiness and maybe its ability to maneuver, and the latter is mainly going to depend on tactics since the only advantage the A10 has is terrain masking and that is very situational.

Sure, the A10 can land with half of a wing left, but it is still out of action even if the pilot makes it. Modern aircraft are really hard to replace - it isn't like they'll be rolling out by the hundreds from assembly lines during a war making the pilots the major limitation. We of course should try to protect the pilots, but that isn't going to make the plane all that much more useful in a war.

All that said I wouldn't look to get rid of them either. The reality is that we don't know what the next serious war will be - better to hedge your bets especially if you can do it with really cheap aircraft. I imagine you could field a lot of aircraft like the A10 with some modern redesigns for far less than you're going to pay to make something in the class of the F35 and F22.

This is only because the practice isn't well established. When you call around for car insurance and find that everybody's default rate is $5k/yr, but they all offer a $4500/yr discount if you accept monitoring, few people will stick with the default. I also wonder how many new Progressive customers don't accept the monitoring - I would imagine that the company would advertise it to anybody who asks for a quote since it lets them give out a lower number which makes them more competitive. Bottom line the consumer is given two rates and has to choose between them.

This will just result in either base prison sentences being shortened, or people being let out early all the same on some other pretense. You can call the label whatever you want, but for various reasons society wants people who behave well in prison for n duration, and people who don't behave well for n+m duration. It would take time for things to change - I don't suggest that prison sentencing is "economically efficient" (I realize I'm being very loose with that term, but I suspect you will get what I'm trying to say without going on about it). Actually, something like prison terms is a pretty poor analogy because there is so much politics involved - voters/officials/etc don't decide on how long to punish a murderer the way they decide on how much to pay for a new car.

To some extent they don't have much choice, unless insurance is compulsory and enforced.

Insurance only works in the absence of knowledge. If I had a magic machine that could predict the locations of every house fire the next day, the fire insurance industry would cease to exist, unless they were allowed to require coverage to be purchased two days in advance without an option to cancel without paying two days premium. Otherwise everybody would drop fire coverage and sign up the day before their house burns down, and insurance companies would basically have to charge the replacement cost of the house for their policy, making it pointless to buy in the first place.

You can't legislate around this sort of thing without mandating universal coverage. If you prevent insurers from discriminating against those purchasing insurance on some basis, then you also need to prevent people from being able to avoid buying insurance on the same basis.

Suppose life insurers can't charge people with diabetes a different rate. Diabetics have a much higher cost to insure, so on average the rate would have to be much higher. A non-diabetic would look at the risk of needing insurance vs the cost of buying it and conclude that insurance isn't worth buying. Now the risk pool changes so insurance companies are ONLY insuring diabetics so they have to REALLY increase the premiums. Previously diabetics were signing up in droves since the rates were cheap, but now there is no incentive for them to buy insurance since it actually reflects real costs. The result of preventing an insurer from charging more for a particular group is that they just charge more for everybody instead, and this is unavoidable as they'll end up there whether they start out that way or not.

Yup, this sounds a bit like Y2K in retrospective. Was money wasted on it because it turned out to be a non-event, or was it a non-event because so much was spent on it?

Always money to do it over, never money to do it right... :)

No argument with that, but in this case Qualcomm isn't publicly accusing anybody of anything. They mailed a private letter to a service provider, and that was it for their part.

Sure - Gentoo Linux. To my knowledge they've never paid a dime to any contributor, other than covering a fairly small number of expenses with the resulting assets owned by the Foundation. SleepyHead is another that I'm aware of - the author accepts donations, but they don't come close to covering his time. Until fairly recently projects like CyanogenMod were purely donation-based and those certainly didn't cover the time involved (and only a few of the contributors actually get donation money).

I have no doubt that people have contributed to many of those projects using time compensated by another employer, but we're talking a very minor share here compared to something like Fedora/etc.

I'd call volunteer projects the norm. Very few projects turn into Apache/Mozilla/etc. For every one of those there are a thousand gtklifes.

The problem with this kind of approach is that it basically punishes kids for the mistakes made by their parents.

It would make far more sense to take care of the people who are actually born and maybe try not to have so many that the state ends up having to pay for.

Maybe parents would be more involved in public education if you had to pre-pay 12 years of tuition in order to be allowed to have a child. :)

Yup - the beauty of actuarial tables is that they contain all those non-politically-correct correlations we're not supposed to talk about. We can hate what is in the tables, but they are cold hard statistics. Certainly they are open to over-interpretation, but the correlations are what they are.