Hacker News Reader

Comment by vineyardmike | Hacker News Reader

Creating a spec that people don’t want, can’t, or can’t afford to adhere is a failure.

I don’t recall ever seeing prior USB devices fail at adherence.

The problem is that USB C is a massive spec with a lot of things that aren’t always needed in it (display, pd, high data speeds etc). You can either design your product to implement a lot more things (eg Macs) or skimp out and cause confusion (Switch). By breaking the standard, people effectively have to buy your charger or cables to know it’ll work. If I have to either buy a high end product or buy cables and chargers only from the original brand, we aren’t really any better off with this.

View on HN

>I don’t recall ever seeing prior USB devices fail at adherence.

Really? You've never seen a power only micro-usb cable (missing 2 bus lines)? I have dozens from all types of devices that use USB as a charging only connector.

This has been a problem with every USB iteration (hard to have the U in USB if you aren't able to handle a wide variety of applications), but I vastly prefer it to the early 00s when you might have 5 different cables to each "specialty" application.

Or maybe Nintendo just decided that they didn't care. I really don't see how any technical specification could be expected to physically prevent everyone in the world from deliberately implementing it partially or incorrectly. I don't doubt that USB-C might be technically complicated and confusing compared to alternatives and predecessors, but I find it very difficult to believe that Nintendo engineering gave it their best shot and simply couldn't manage to do it correctly.

> I find it very difficult to believe that Nintendo engineering gave it their best shot and simply couldn't manage to do it correctly.

I'm sure that's not it. I'm sure it was

"Hey product manager. We can launch in 6 weeks if this port accepts power and connects to a dock, or 36 weeks if we need to conform to standards in section 17, subsection 5 part 41.2B of standard 1120 on European code B7"

"what does <standards> get us"

"Nothing, we don't need those features. Also it'll add $25 to the bill of materials"

"ignore it then, we want to launch sooner and increase our margin."

Perhaps so, but how is that something you can blame the spec for? Any remotely non-trivial spec will be easier to implement partially than to implement completely.

Maybe the spec should be more “trivial”. I like USB C over something like micro usb but micro usb is easy enough to use as a power source in my high school electronics class but i don’t think usb c would be.

Maybe apple had it right originally to have “thunderbolt” as a super powerful all-in-one cable (except add power) and we needed a simple “reversible micro usb” for everything else.

Sure we are, if they don't conform to the spec then they can't get a CE mark; which then means it can't be sold in the EU and the device misses out on a 350~400M person market.

I think any protocol that wants to implement a wide range of features will be complex.

Afaik if you don't need these you can just implement the parts you need e.g. charging.

Reply on news.ycombinator.com