I doubt a magnet that small would cause any problems. I'm not sure about the new MacBooks Pros, but the old ones had an embedded magnet that I think was used to detect the screen closing.
In Lincoln, NE you can get a one bedroom apartment with utilities included for $425 and if you are happy with 15' x 20' one room apartment you can get one for $275. There is a college here so finding interns is fairly easy and there are lots of small startups. However, once you start growing you might have trouble hiring enough people since there just aren't as many people who want to be in Lincoln as there are people who want to be in Silicon Valley or the east coast. I'm not very social so I don't know how good the socializing is here.
If they wanted an Objective-C web framework it seems like they would have been better off improving GNUstep Web. From looking at the site I don't see what makes this special other than using Cocoa. It might have some compelling features that GNUstep Web doesn't, but they don't make it easy to find out what they are.
Getting the power to the US from a large third world nation is probably not as cheap as easy as building them in the US. There is also lots of land in Arizona that isn't useful for very many things.
Why do you think that Microsoft products are necessary to sell to large businesses? There might not be any large businesses that don't have any Microsoft products, but I've seen several successful sales to very large business by small companies without Microsoft products.
That seems like it could easily end up being very slow. The OS doesn't know what data is going to be needed and it could page out the wrong data. It's probably worth trying, but it doesn't seem like a masterful solution.
Sitting next to someone tapping away on their phone during a movie is distracting and saying that other people are doing it as well is just a childish excuse.
The deployment may be cross platform, but to develop with Cocotron you need a Mac. I didn't see the reasoning behind them choosing Cocotron explained on the project page, but it seems like if they really wanted it to be cross platform they should have used GNUstep. It would also be nice to have a comparison between this and GNUstepWeb.
That seems like a much better comparison. However, what it leaves out is the network usage. Maybe there aren't as many people using one carrier's towers in those other places or maybe the usage isn't as heavy. So while we know that AT&T's network can't handle the load we don't know if it's because of an inferior network or because of higher usage. I would be interested in seeing those kind of statistics compared for well performing and poorly performing networks, but I don't know if that information will ever be available.
Being sparsely populated makes it much easier to run a stable network, so that's not a very good comparison. It's not that AT&T's network is bad everywhere, but places with lots of people in a small area cause problems. Where I am the service is excellent. I've never had a dropped call here and the data speed is always high.
If you have 100% code coverage you don't necessarily have 100% path coverage, so there could still be errors that a statically typed language would catch at compile time. Realistically, most programs aren't going to have extensive test suits that get anywhere near 100% code coverage, so static typing becomes even more valuable. Also, running a full test suite for every build would take much more time that running a type checker.
C is not a bad language for humans. It's simple enough to be able to see how the machine will execute it but not low level enough that you have to manually manage everything, like assembly. It's not always the best choice, but for some things it is. It's also good to know so that you understand how how computers work instead of assuming that higher level languages are magical.