in the release notes of new software:

(Are you listening, Mozilla?)

[AustralisSucks]

The thermostat runs the heater based exclusively on the outside temperature. This is laughably called "weather controlled heating". The basic idea is that the building loses heat to the environment based on the temperature difference. So if you know the outside temperature and the rate of heat loss, you should be able to get a constant temperature inside, no?
Of course there are big problems with this:
- The placement of the temperature sensor (yes, just one) is critical. If it's in the shade, it'll miss the effects of insolation.
- It assumes the building is homogenous when in fact, the sun-facing side is always too hot and the shaded sides are too cold.
- Fine control depends on thermostatic taps on each radiator, but thanks to being right on top of the heat source, these generally don't work.
- The heating curve programmed into the thermostat must match the actual characteristics of the building, but no HVAC installer I know ever calculates this. Instead the curve has to be adjusted by trial and error. I know at least one installation where the interior temp is often off by +5 C despite multiple attempts at adjusting the curve.
All of this idiocy could be prevented by placing a temperature sensor inside the building. Even a single one located at the thermostat control panel would work. But no, that would be too simple.

It'd be interesting to see how they plan to do this. The main obstacle to mass production using CFRP (or any fiber-reinforced plastics) is that it takes much longer to put fibers in a mould, impregnate them and have the mixture dry to the point where it can be removed from the mould, than it would to stamp a sheet of aluminium into shape.

"This month, we set up a new system for department X that helps them accomplish task Y in 50% less time"

I tried something similar with some audio cassettes a few years ago, and found that I was too late: the tape had begun to stick together, and required more power than my high-end Denon tape deck could muster to play back. Rewinding didn't work either, as there's a tape tension sensor that shuts down the motor if it gets overloaded.

The battery in my Casio lasts 5 years.

I'd say that needing regular adjustments (every other month at least) falls under "doesn't function all that well".

Look up Philips Wake-up Light. The same idea, but with gradually brightening light instead of instant sunshine.

A long time ago, I used to volunteer for a local radio station. They had a couple of Studer PR-99 open-reel tape recorders. These were used pretty hard, all day long. We used them so much that the recording and playback heads developed visible grooves from all the tape that spooled past them.
They were incredibly overbuilt. Weighed a ton, they had 3 electric motors that looked like they could start a car. All the circuit boards were slotted in for easy maintenance. They never failed, despite the abuse they went through.

Using seawater in the secondary cooling loop makes maintaining the cooling system a nightmare because seawater's rather corrosive.

Use a penlite instead with much more capacity for 1/20 the price.

If engineers were left to design a parking app, they'd make it work for all forms of paid parking in at least one country. Instead, we're saddled with a fragmented market.

Taking homeopathy to its logical conclusion: plain ordinary tap water contains homeopathic concentrations of every substance known to man. Ergo: water should cure everything!

Yes, this would make things simpler. The French have done this (PDF link), using one standard reactor design wherever possible. IIRC the American method was to use some standard components, but allow the architect responsible for the plant to make lots of changes (e.g. the piping between the standard components is different at each plant).

In a front-wheel-drive car, left-foot braking can help stabilize the rear: you get a net braking force on the rear axle, and some drive to the front.