Cal

Doorways: go over the frame, not under the door, and it is not close.

Under a door the cable gets stood on every time somebody walks through, and a door bottom is a shear point — it does not crush the cable so much as work it back and forth until the conductors inside fatigue. That failure is horrible to diagnose because it is intermittent long before it is dead.

Over the frame the cable is out of everyone's way and the architrave usually gives you a natural shadow line to hide it in. Run up the side of the frame, across the top, down the other side. It sounds like more work and it is about ten extra minutes.

If you genuinely cannot go over, there is flat under-carpet cable designed for exactly this, but it is a different product and you want it by design rather than by squashing a round one.

Washing: it does something, and it is not what people hope.

Rinsing removes spores sitting on the surface, which is worth doing and will slow down how quickly the rest of the pack goes over. It does nothing whatsoever about a network already growing inside a fruit, because you cannot rinse the inside of a tomato.

One genuine trap: wash them when you are about to use them, not when you get home. Washing leaves moisture in the stem scar and in any tiny crack, and that moisture is the thing mould was waiting for. A washed and re-boxed punnet goes over noticeably faster than an unwashed one.

The storage version of this: get them out of the sealed plastic. Those packs trap the moisture the fruit gives off, which is why the bottom layer is always the one that goes first.

On your third question, the thing that actually fixed this for me: stop buying fasteners for the job and start owning a range.

A sorted assortment box of M3 to M8 in coarse pitch, in a few lengths, costs less than three emergency trips and removes the entire failure mode for maybe 90% of household jobs. Keep the removed screw taped to the box lid until the job is done so you can compare directly.

The remaining 10% is fine pitch, imperial hiding in old or American equipment, and self-tapping screws that are a different system altogether. Those you take the old one to the shop for.

The number is the nominal outside diameter of the thread, in millimetres. An M5 screw has a thread that measures roughly 5mm across the outside of the ridges. That is it — it says nothing about the head, the length, the drive type or the strength.

So M3 is about 3mm across the threads, M4 about 4mm, M5 about 5mm. If you put callipers across the widest part of the threaded section you will get a number very slightly under the nominal, because the tops of the threads are truncated. 4.8mm on the callipers is an M5.

The reason two M5 bolts look nothing alike is that everything else is quoted separately: M5 × 20 is 5mm diameter, 20mm long under the head, and the head could be hex, socket cap, button, countersunk or flanged without changing either number.

Your binding bolt is the third dimension nobody prints on the front of the packet, which is what your friend meant.

The visible losses, in the order most people notice them:

Subject separation. A small sensor with a wide lens keeps almost everything in focus. That is why portrait mode exists, and why it still fails at hair, glasses frames and anything with a gap in it — it is estimating a depth map rather than measuring light. On a normal photo of a person in front of a busy background, that is the difference you feel without being able to name it.

Reach. Digital zoom past the longest real lens is upscaling. A phone at 10x and a modest telephoto lens are not comparable, and this one shows up constantly in ordinary use — children on a stage, a bird, anything across a street.

Motion. Merging frames assumes the scene held still. A moving subject in dim light is where phone shots go soft or smeary, because the frames did not line up. A single fast exposure from a bigger sensor just works.

All three are ordinary photos, not test charts.

There is a technique fix for the sticking specifically, and it is heat, but not in the direction people assume.

Ridges have far less metal in contact with the food than a flat base, so they dump their heat into the first thing that touches them and recover slowly. If you preheat to the same temperature you use for the skillet, the ridge tops crash the moment the food lands, and food that touches metal below searing temperature bonds to it. That is your sticking.

So: preheat longer and hotter than the skillet, and give it real time — ten minutes on a medium flame, not three. Cast iron ridges are thick and heat unevenly across the pan. Then oil the food, not the pan, because oil poured into a ridged pan runs straight into the channels where it is no use and smokes.

And do not move it. If it is stuck, it is not ready; a properly seared ridge line releases on its own.