Hacker News Reader

Comment by code-faster | Hacker News Reader

There has to be a post-mortem on this.

My take is that the standard is too lenient and the cables aren't color coded.

Lenient standards mean that "usb-c" could support 1W or 10W and your phone plays footsy to figure out how high it can go. Pick a standard, pick a level, don't exceed the standard even if you can.

Color coding? Yup. Colors. A human must be able to visually determine what specs a cable matches and doesn't match. Before, different standards had different ports, so color didn't matter. But usb c wants to be the same port for multiple standards. Fine. Color code it. It can even be a small dot by the plug, as long as it doesn't wear off.

View on HN

> Color coding? Yup. Colors. A human must be able to visually determine what specs a cable matches and doesn't match.

Shapes (of, e.g., logos) are much better than colors as a universal label for this, though colors as an additional market are fine. Colors alone are a problem because of common defects of color perception.

yeah a 2d logo on the connector or a cable pattern is better, good call.

How about keyed connectors and sockets? Why have incompatible cables with identical form factors?

My mind is blown.

Everything (recently) old is new again.

Changing a cable that can charge my phone (but slowly) into a cable that can't physically fit into my phone... probably isn't an improvement?

The connectors could be keyed such that backwards-compatible functionality also has backwards-compatible connectors.

Can you explain how this would be different from the current situation? When both cables and devices are usually backwards-compatible, what do you even key?

The main thing I can think of to key out is charging-only cables, but those aren't even supposed to exist as far as I know. Cables are all supposed to have at least the USB 2.0 wires.

USB 3.0 B-Type and USB 3.0 Micro B will allow USB 1.1/2.0 cables to connect to the respective ports, but at lower speeds.

Last two on https://www.cablestogo.com/learning/connector-guides/usb

So you want keying that prevents you from using a too-good cable, like with micro? Cables either have to match the port speed, or be worse?

It's less about being prevented from using a cable thats "too good", and more about knowing what a given cable supports without having to plug it in. If a cable and both devices had this hypothetical USB-C+Thunderbolt keyed connector, then it would be clear by visual inspection that all 3 components necessary support that level of connection. Instead we have the confusing situation that we're in where the only real answer is to try it.

Even at $70, the cable is likely the cheapest component of the three and I would rather know that I've got something that works rather than take two trips to the store.

Yes, on the off chance that you're teleported somewhere and end up with a laptop, 4k monitor, and the slow cable, you won't be able to get 4k because of the cable, but imo that trade-off would have been worth it compared to the current confusion.

But USB-C is universally-comaptible with capability detection and least-common-denominator functionality, so you are back to one connector with no actual keying if you do that.

I’d rather be able to charge my devices slowly than not at all. To be honest, I really don’t care about fast charging. I’m carrying around battery packs either way.

Exactly, the main complaint with using the wrong cable is the charging is only at regular usb speeds, or the data transfers at usb 2.0 rates. That's way better than nothing, and usually doesn't matter too much.

We’re talking about a standard that was invented primarily for laptops. That is, it was invented to replace the wide variety of single purpose ports with a few universal ones. Keying the connectors defeats that purpose since you now need different ports again.

I know people love to complain endlessly about Apple taking away all of the single purpose ports on MacBooks and replacing them with USB-C. The problem the article is describing is a problem of counterfeit and mislabeled or unlabeled cables as well as improperly designed devices. For most users this isn't a problem and the ability to charge the laptop and plug in any device in any port is very convenient.

In theory, that would be convenient. In practice, you get stuff like this:

https://apple.stackexchange.com/questions/363337/how-to-find...

TL;DR: if your mac has high cpu load, try using the other side port for charging.

Note that this is irrespective of the mess the article talks about, which just exacerbates the problem.

In USB-C land, some (Endpoint-to-endpoint) connections are definitely a lot more equal than others.

> How about keyed connectors and sockets?

The only cable I can think of that had keying for capabilities was DVI, and that turned out awfully. So many times you had cables that could work, but the keying rejected them.

Keying is much worse than simply having cables of different speeds.

> Why have incompatible cables with identical form factors?

They're not incompatible.

Agreed. If a cable plugs in it should "work", and all usb c cables "work". I can use a crappy usb c cable to charge my laptop. The issue is that how "well" the cable works varies wildly and is not visually determinable. I can use this cable to charge my laptop...but how long will it take? That's the issue

They aren't incompatible, that's rather the point with USB-C. Any combination of compliant cable and socket works with the capabilities of the least capable component in each dimension. Everything just works, though some combinations work better for certain tasks.

white with a blue stripe or blue with a white stripe?

I vote shapes: wider = bigger pipe, the analogy works.

Plus I would delight in saying, "Yes, but higher power means bigger electrons..." to non-engineers.

Reply on news.ycombinator.com