I wouldn't count on it taking any job from a CNC machine. it's a MIG welding machine and it's resolution/accuracy should be well over 1mm. It won't deal with a lot of metal also.
No, it's not the one size fits all.
It's very useful and interesting, allowing complex geometries more easily and I think local shops will have one once the techs (his and other techs for metal 3D printing) are mature.

I think you never had to use cygwin...
Build time on modern computer could be an issue sometimes, but your statement that the performance of the binary is more important keeps valid.
Not that the benchmark of the article makes any sense...

Those things are usually installed by engineers with very little knowledge/concern about security. In my field there's an urge for bluetooth connectivity for the industrial equipments, with all the security nightmares bluetooth poses on accessing a device. Wi-fi could be used with a much better security model, but it's considered too complicated...

I've made embedded programming in C++, in a rather efficient way. It makes a lot of things easier, but it's much harder to find someone that knows what that thing is doing. If you handle a task to an experienced C++ programmer and say no use of new, delete and exception handling are allowed, probably the task won't get accomplished unless it's very simple.
I used to advocate the use o C++ on embedded, but gave up. In a larger environment with programmers of different experience levels, C is much easier to get running.

Unfortunately, not even that. They'll find an error in calculations, another doomsday theory or whatever....

Hit the nail.
If you have at least a little experience, you've already at least once gone through the codebase and asked who's the idiot that wrote that, checked the answer on the version control system to find that the idiot is yourself.
On the other hand, sometimes the legacy code is so much patched that it's easier and faster to rewrite it. If this code is that much patched that you can't even follow the code flow anymore, bugs left are only the scary ones anyway. Rewriting that code is a high risk task, since many errors already solved will be done again, but may still be worth it.

Li-ion batteries (laptop batteries) have really small changes in voltage output until very near the end of the charge. There's indeed a very precise integrator calculating the estimated remaining energy on the battery by measuring current and voltage.

No, redundancy means excess capacity.
Diversity means spreading the capacity through various locations/processes/etc. In this case we're talking about geographical diversity and the difference is simply distribution of the fabs around the world. If TSMC had its 15 factories spread over various locations, a single event couldn't shut down 100% of it's capacity.
The last year's earthquake in Japan shot down 1/3 of the Renesas fabs. It was enough to make a huge problem, stop some auto manufacturers, but it kept running. Had Renesas all it's fabs on North Japan, it would be 100% down. The fabs were not redundant, as one could not cover for the lack of other.

It's about diversity, no redundancy.
For semiconductor manufacturing, for example. There's quite a few plants around the world, not that much of small process nodes. If a major earthquake hits Taiwan, shutting down TSMC and UMC factories, you'll notice the effect all around the world pretty fast.
This http://eetimes.com/ContentEETimes/Images/120514_icInsights_micron_800.png should give an idea.

That's because there's no "sad but true" mod.

Having compiled the same code for ARM, x86 and x86-64, with the same compiler, I dare saying that the ARM code is much smaller.
Compiles using -Os and GCC for all cores.
Embedded has code size as a limiting factor for quite some time now...

P.S. before saying that a sample size of one isn't evidence, I develop firmware for ARM and some other arch e frequently compile it on the PC for testing purposes.

So far the problem only got worse, so all of these could very well be right on the supposition that it's now irreversible.

No, they are not.
Means for implementation by an expert should be provided. The C++ code would be just an example of implementation. Try coding a MPEG encoder in Lisp and see if you can patent it.

Indeed. Since FF4 I can't use online banking with FF. Now with FF5 Acrobat Reader locks and fails every now and then. Befor you put the fault at adobe, it doesn't on FF3 or IE. Basically FF is becoming useless for me very, very fast. And I'm the ordinary user. On the enterprise level here it's already forbidden.