Agreed.

Can anybody explain what's wrong with just using Oauth 1?

I used to administer a Sparcstation 10 being used as a server in a lab. The only time I ever needed to reboot it was for an OS upgrade or a power failure, and those happened more than a year apart typically.

FWIW, it's becoming fashionable in Silicon Valley to use solar panels as shade structures in parking lots. The cars get much-needed shelter, and solar panels get installed without covering up anything like lawns or gardens. It's a complete win-win.

This is key. Unless your surface area is limited (space craft, vehicles), it's not efficiency that matters, but cost per watt of capacity.

Make solar cheaper per watt than coal plants (we're getting close now), and then watch all the rooftops in the country get covered with solar panels.

Even if all the rooftops combined aren't enough to produce *all* our needs, every 300MW of solar power is one coal plant shut down, and 2400 tons less CO2 produced. Per day.

I know someone with a Rav-4 which she charges from solar panels on the roof of her house.

However, unless there's a serious revolution in battery technology, I don't think the electric car is ever going to be practical.

Likewise, solar panels don't work at night, under trees, or when it's cloudy.

None of those are arguments against developing solar technology. Or wind power. While neither of these can ever totally replace fossil fuel power or nuclear, they make excellent supplements. Solar power is at its peak at the same time demand is at its peak. And every kWh of solar power represents 1-2 lbs of CO2 not released into the atmosphere.

No I feel you should comment it out for one version, or one iteration.

This is exactly what I do. Sometimes I leave it for a couple of versions, but basically you want it to still be there long enough to be sure that deleting it was the right thing, then get rid of it to clean up your code. Leave a good comment in the version history and you're done.

One of the Eureka's sister ships (or was it the Eureka itself?) was used in Africa for exactly this kind of research for a while.

OK, speaking of one who's actually taken dirigible flying lessons, I have a couple of points to make:

Other posters are right: propellers are just little airfoils.

The ceiling of a prop plane is a combination of three factors: thin air limiting the lift of the wings, thin air limiting the thrust of the prop, and lack of oxygen to the engine. Superchargers can help with the oxygen problem, and longer wings and/or higher airspeed will help with the lift problem, but there's not much you can do about the prop.

Airships have altitude limitations too, even worse than airplanes. Every airship contains air bladders called "ballonets" which displace some of the lifting gas. As the airship gains altitude, the ballonets are deflated to make room for the expanding lift gas. Once the ballonets are completely empty, the airship is at its maximum altitude, beyond which it can't rise without venting and losing lift gas.

Airships are *not* "extremely efficient at sending hundreds of tourists plunging to a spectacular death". The Hindenburg caught fire a hundred feet in the air, and most people on board still walked away. You can't say that about most aircraft. We think of airships as dangerous because the Hindenburg disaster happened in the relatively early days of aviation, and the disaster was broadcast live, searing it into the collective consciousness.

The Hindenburg itself was a very safe design. The disaster happened because they screwed up and used highly flammable paint on the skin. If they hadn't done that, things would be very different today.

All that said, there are a number of factors that will keep airships from ever coming back.

First, the cost of Helium is going through the roof. This is essentially what killed Airship Ventures. You could make a reasonably safe airship using hydrogen, but nobody would be willing to fly it. This might work for cargo transport, but not for passengers.

Second, they're slow. Third, they don't operate in high winds.

Flying one was one of the most seriously awesome fun things I have ever done, but I have no illusions that they'll ever be a practical means of transportation again.

Disagree. I used them all through college. They don't have to be held perpendicular, but they do require a light touch, and as I mentioned above, the finer ones were a maintenance nightmare.

Yes, they're intended as drafting pens, but a 00 or 000 makes an excellent writing instrument.

I used Rapidograph pens all through college, and still have several of them. The "sweet spot" for an engineer's notebook was the 000 size, although I often carried both the 00 and 0000 sizes instead. I knew one guy who carried the 000000 size, but the lines were almost too fine to see, and they tended to cut up the page.

Anything finer than 00 was a maintenance nightmare, though. They clogged easily, and the wire that ran up the tip was likely to bend during cleaning, ruining the tip for good.

Nowadays, I just write with a traditional fountain pen, with as fine a tip as I can get. These suffer from problem #3, especially if the page gets wet later, but they write a beautiful line.

Other than that, I have a "Sanford Uni-Ball ONYX fine" on my desk that I use when I've mis-placed my fountain pen, and it also writes a superb line. It also will make carbon copies, which neither a Rapidograph nor a fountain pen can do.

I saw that episode. Henry hooks up with a hot young thing that seems too good to be true, and then she runs around the camp hitting on all the other officers too, stirring up no end of trouble.

Assuming this story is even true, the only morals to be gleaned are: "If it seems too good to be true, it probably is" and "don't stick it in the crazy". There is nothing to be learned here about relationships and technology.

Because the underlying toolkit libraries are in a constant state of flux, with each version being incompatible with the previous, applications are all subject to bit-rot. An app that worked in 2008 will very likely not work today unless the author went to the trouble of porting it to the new toolkits. This is true for both gtk and qt. And don't even talk about motif and olit.

But you need the basics.

In my entire career, I think I've only used math once that I hadn't learned in high school (differential equations for a fluid flow simulator).

I've used algebra, matrices, geometry, and trigonometry on a pretty regular basis for the bulk of my career. A lot of that career included computer graphics, which uses matrices and trigonometry heavily.

For years, service providers have been beating up their customers to get them to use secure passwords, but time after time, it turns out that the service providers are the worst security offenders.

What is it going to take to get the services to take security seriously?

It's not that hard: Build a dedicated authentication server. Account names and passwords (preferably hashed) are stored there, and NOT in any other database on any other server owned by the service. The authentication server acts as a near black box, accepting credentials and returning a simple yes/no answer. Only a very few employees have access to the authentication server. Naturally, the server itself sits inside the DMZ, inaccessible from the outside world.

It might not be perfect, but it would have stopped all of the major password breaches I've ever heard of.

Great. I was starting to learn QT3, not realizing it was already obsolete. And QT4 has what, three years left before it starts breaking?