The laptop power supply became a generic part in April. It is a good outcome and a small one.

There are four USB-C cables in the drawer under my desk and I cannot tell them apart. Two arrived with phones, one with a dock, one from an airport shop at a price I would rather not have paid. Same length, same black, same connector at both ends. One of them will charge a laptop at full power. I do not know which one, and neither the cables nor the laptop will tell me.

That is the entire complaint, and it is worth saying at the outset that it is a complaint about a situation that has got better.

What changed, in one sentence

Since the spring, a laptop placed on the European market has to accept charge over USB-C, which means the power supply is no longer a part any one manufacturer owns.

The date sat in a directive from 2022, not in a keynote. It set one application date at the end of 2024 for phones, tablets, headphones and a list of smaller categories, and a later one this April for laptops, which got the extra time because the power levels and the enclosures are harder. I am paraphrasing from the published text rather than the consolidated version, which is why the source note asks for both dates to be checked. Nothing in what follows turns on the day.

The cost of it landed almost entirely on industrial design, and it landed years ago. A charging port is a hole in a casing decided long before the casing is cut. By the time the deadline arrived, the argument had been over for two product cycles.

The part that now outlives the machine

Here is what that does on a five-year view, which is the only view worth taking.

A laptop power supply used to fail on roughly the same schedule as the laptop — the cable frays where it leaves the brick, always at the same angle — and it was replaced, if at all, by an object with a part number, a single supplier and a price set without competition. The part outlived the model in the warehouse sense but not in the useful sense. You could not put it on anything else, and nothing else would go on it.

Now the failing part is a cable that costs less than lunch and fits four other things in the house. That is a real improvement and it is the sort that shows up in year three, quietly, as a problem you did not have. It also inverts the lifespan: for the first time the charger in the bag is likely to be older than the device plugged into it, and to survive it.

The unbundling half of the rule pushes the same way. A buyer can take the device without a charger, and the box has to say which arrangement is on offer, so the default drifts from one power supply per device towards one per person. Fewer bricks in the drawer. Slightly more pressure on the ones that remain.

The connector is standard. The cable is not, and the cable is the part you actually end up owning.

The label answers the device's question, not the drawer's

The device now carries information about what it needs: the power it draws, and whether it speaks the fast-charging protocol. That is useful at the point of sale, which is where the rule was aiming.

It is not useful at the point of failure. The cable is where the ambiguity went. A plain USB-C cable carries three amps; carrying five requires an identifying chip in the connector, which is invisible, and which the outside of the cable usually does not mention. There is a certification logo programme, and in my experience the logo is missing far more often than it is present. So the drawer contains four objects that are physically identical and electrically not, and the laptop's response to the wrong one is not an error but a slower number.

Which is the worst possible failure mode for a person to debug: it works, it is merely worse, and nothing anywhere says so.

I could not establish whether any labelling duty attaches to a cable sold on its own. The texts I have read address the device and the charger that comes with it. If such a duty exists somewhere I have not looked, this section is wrong, and I would rather be corrected than hedge it.

This is not a turning point in anything

The temptation with a rule that worked is to write it up as though it proved something larger about industrial policy. It does not. A standard connector removes a category of waste and a category of annoyance, and that is the whole of its ambition. Nobody's five-year device economics are transformed by it.

It is worth noticing, though, what kind of rule succeeded. This one specified a physical interface and a date, and left everything else alone. It did not ask anyone to publish a commitment, maintain a network, or hold inventory — the three things that quietly decide whether hardware survives its fourth year, and the three things that keep going wrong. A port is easy to mandate because a port is checkable by a person holding the device. Most of what matters about longevity is not.

What breaks next

Not the port. The failure mode of the next five years is a laptop charging at half speed off a cable that looks exactly like the one that worked yesterday, in a hotel room, at eleven at night, with no way to tell them apart and no error message to search for.

The connector problem is solved and it stayed solved. The problem of knowing what you are holding is not, and there is no directive drafted about that, because writing on a cable is not the sort of thing anyone legislates until it has annoyed enough people for long enough.