Because all sorts of services won't talk to you if you are on a spam blacklist. And pretty much all dynamic address pools are on such blacklists.

The GPL3 does indeed state that. But isn't the purpose of section 7 only to allow the copyright holder to work around the fact that the header of the license states that changing it is not allowed? Section 7 explicitly mentions that you have to have permission from the copyright holder to add additional clauses. You may thus add restrictions to the license which are not counter to the spirit of the vanilla GPL3.

But the question is, how does this affect compatibility between "GPL3 with additional restrictions" and just "GPL3"? Since without permission from the copyright holder you cannot add or remove additional restrictions, you may not use the licenses interchangeably. Because the additional restrictions are not present in the vanilla GPL3 (which does not allow additional restrictions unless you are the copyright holder and these restrictions fall under section 7), if you combine GPL3 code and "GPL3 with restrictions" code, it seems to me that there is no possible way to satisfy the terms of both licenses simultaneously.

So the way I see it, although GPL3 allows you to add these terms, by doing so you make your license GPL3-incompatible. If this is the case then it is troubling, and I would welcome any clarification from someone who knows the details behind this.

It seems that the licensing is a mess.

The header in the source files state that the code is GPLv3 or any later version, with additional clauses added.

However, it seems that it is only possible to apply these additional terms to GPL version 3 exactly (and not any later version):

These additional terms seem to be just disclaimers of liability and an indemnity clause, but it is entirely possible that they make the source GPL-incompatible, which, if true, would be a huge disappointment.

So not only is the license not self-consistent, it is likely also GPL-incompatible. The additional terms may further make the license non-free, and definitely non-DFSG-compliant. Thanks go to the corporate lawyers who have turned Carmack's good intentions into an abomination. I hope that they can re-release this under saner terms.

I can never tell if these words in quotes are meant to be taken as literary citations or an indication of sarcasm. I think this 'style' of writing should be 'retired'.

We don't need any breakthroughs; we already have good plans to get to Mars with existing technology. Have you heard of Mars Direct? It's a plan for routine manned flights to Mars accomplished with existing chemical rockets (actually the Shuttle launch stack), artificial gravity induced by spinning the spacecraft linked with a tether to an empty rocket stage around the common centre of mass, and using chemically propelled cars (why would we use electric if we can have something better?). If we had wanted to, we would have sent people to Mars by now. All the things you mention are completely unnecessary.

This is exactly what NASA should not be doing. If you want to know why, you should read the 2003 testimony of Zubrin on the future of NASA.

Not all ARM CPUs have division. I know for a fact that an ARM7TDMI doesn't.

Is there even one mature language that runs on Parrot? (not brainfuck, etc.) They seem to be in perpetual-development mode.

No, the main advantage of objects is convenient interfaces.

Not for unsigned numbers it isn't.

To my knowledge, there is still no solution. But if you read it carefully, the article does leave one possibility open: assemble a large heat shield in space and use it to decelerate through atmospheric drag on Mars before opening a parachute. We don't currently have the capability to do it, but it seems technically feasible.

...and disabled.

No, as a matter of fact it was a rip-off of a novel by Tom Clancy (read the last paragraph of the plot summary).

That's GNU/Kernel.dll, mister!

Speak for yourself, I want my grave to be made of glass so I can look up girls' dresses after I die.