Peak oil is a mathematical certainty. One day we will be no longer using oil. This means it is mathematically certain at some point there will be a global maximum to the oil production function.

I have dozens of LEDs and had so far only one failure in 4 years (I thought I had a new failure the other day, but there's a bad connection in the lighting circuit - the lamp works fine when put in another fixture), so "rarely last more than 4 years" seems unlikely at this stage.

The colour rendering is good enough, I replaced my kitchen lights with LEDs about 4 years ago and have not noticed any issues with colour rendering. A turn-on delay of 100ms is functionally instant (I can't imagine any case in my home where 100ms for a light to turn on will be significant - and tungsten bulbs will require at least this long to warm up and glow at full brightness), and they are full brightness immediately (unlike GU10 CFLs that take 4-5 minutes to warm up). The EM emissions are utterly trivial. At the frequencies the current regulation electronics works at, there is nothing on the internal circuit board that is long enough to act as anything remotely close to an efficient antenna, and none of the lights I have bought are "dodgy" - they have all been tested for conformance to EM emission standards. They do not interfere with any radio equipment I have.

My bike light is a seriously bright LED headlight (and tail light). The headlight is as bright as a car's headlight, and I ride on quite a few unlit roads. Once I've been out of streetlights for a while in the very white pool of light the LED headlight gives me, whenever a car comes the other way, their lights look as orange as sodium lights. Normally I'd see the lights of oncoming cars as more or less white.

The system as a whole is far from 100% efficient. Once you add transmission losses and generation losses, electric resistive heating will be probably below 25% for the system as a whole, as most of the heat will be getting made outside of the building that's being heated.

Gas heating would be far more efficient as a system, since the losses in shipping the gas to your house to burn will be less than shipping the gas to the power station, burning it there to make heat, turning that into electricity, transmitting the electricity, then turning what's left back to heat in the building.

That's surface temperature. This ocean is deep under the surface.

So the follow-up question would be: if it's deep under the surface, how will sunlight get there?

The answer is that sunlight isn't what's needed, it's the right amount of energy that's needed. The energy can come from a lot of other places. Tidal forces for example can heat the interior of the moon, radioactive decay can heat the core of a moon etc. so there may be quite a bit of subsurface energy. For example, if you look at the bottom of the oceans on Earth where there is no sunlight, there are oases of life around volcanic vents on the ocean floor (which are spewing all kinds of useful chemicals into the environment). So while the surface might be cold and lifeless, it's possible that there are significant amounts of subsurface liquid water at a temperature that's compatible with life of some description.

I joke that if PGP stands for "Pretty Good Privacy", PHP stands for "Pretty Hopeless Privacy". The language practically encourages people to write code that suffers from SQL injection exploits.

Science doesn't have proof. Only mathematics has proofs.

Science has theories and evidence to support those theories (or evidence to falsify the theory).

"We don't like change" is the rough translation.

Those icons look no uglier than the old ones.

I feel that any developer who calls themselves a "rockstar developer" is probably suffering a severe case of the Dunning-Kruger effect.

All the really competent developers I've ever known have had anything but "rockstar" like qualities. They generally don't boast, they are generally frugal, they are generally the exact opposite of a rockstar.

Building a simple SoC that works on an FPGA is sufficient. You're getting low enough level by writing HDL to make a simple computer on a FPGA.

And this is why such services actually exist. For example, in the nearest town (pop. ~30,000) there are two shops that will do repairs on things like iPhones/Android phones (the usual stuff - repairing broken screens, replacing dead batteries, removing the SIM lock from any locked phones, replacing home buttons that have stopped working and the usual other wear-and-tear failures that smartphones suffer over time).

I lived in the US for a few years. We all knew it was the richest country in the world (and much richer than the country I'm from) but I was astonished by how common obvious poverty was. I thought our inner cities were bad, but I'd never seen things like trailer parks and some of the small towns in the south that look like they belong in the third world.

We've had chip&pin here now for over a decade, and people still forget their cards.

However: in nearly every system you can put your card in while the cashier is still ringing up your goods, you don't have to wait for the total to come out. When the total does come out the wait for the transaction to complete after entering the pin is normally well under a second on any remotely modern system.

Most European countries forbid bribery of foreign officials. As an example there's a recent case in the UK where two company directors have been convicted and sent to prison for bribing foreign officials.

Some time ago when I was working for IBM, I did have a POWER workstation (back in the days they were called the Risc System 6000). Our entire department had them.