Hacker News Reader

Comment by bluGill | Hacker News Reader

A large number of standards are what one company did on their own and then made available to everybody via the likes of ANSI/ECMA/ISO. There are a number of reasons a company would want to do this.

View on HN

This is basically how USB C became a thing. A group of engineers at Google created the connector design and then gave it to the USB standards body to use as the next gen.

Citation, please. More bluntly: I simply don't believe you.

Google has absolutely none of the expertise required to put such a connector together. It requires a very specific set of skills in chip design, signal integrity, and connector design which Google has none of.

There was a similar rumor making the rounds about Apple, spurred by Gruber.

But 9to5 Mac did some sleuthing and it seems that, in the case of Apple, it’s not entirely truthful: https://9to5mac.com/2015/03/14/apple-invent-usb-type-c/

Google also doesn’t seem to have fulfilled a bigger role than Apple did.

Apple is, at least, far more plausible as they did design a connector in the same timeframe.

I have no idea if Google had anything to do with USB C but "design" doesn't mean "build schematics for". It could be that they just had a few specs they liked and thought would make a good connector and did a rough proposal of form factor and specs.

That's not useful. Here's my diagram for a new connector -> .

Connector design like this is a delicate balance between signal integrity (wants biggger) and mechanical integrity (wants bigger) and size (always smaller).

As for signal integrity: A lot of the signal integrity was papered over by having complex interface chips. This is why there is so much training and negotiation in USB-C.

As for mechanicals: The whole point of USB-C was to take failure-mode data from the previous generations and design a connector that avoided those. USB-C, in spite of how many people bitch about it, was designed so that the the most probable failure modes (which they learned from prior things like mini and micro USB) occur in the cable--ie the replaceable part.

None of this design expertise is inside Google.

In practice none of that is true, the cables hold up fine and the ports wear out. Lightning was a FAR superior design and form factor to USB-C, where what you’re suggesting was actually the case.

I have never encountered an iPhone with a dead lightning port, it’s always the cable that wears out. There’s tons of laptops and android phones with worn out USB-C ports

Go look at the number of engineers that were on USB-C. Apple, not Google, contributed 18 of 79 named engineers listed on the connector certification project, or under 23%.

Intel had the most, as well as the editor position.

I think those 79 engineers should have paid more attention to compatibility and longevity. The cable should have been tested in a dirt tumbler. There is no way a cable should have 32 pins without some kind of built in self test of connectivity on all the pins every time you plug it in. Pins should be reassignable so that when 6 of the 32 pins are dirty/broken the cable still works at a slightly reduced speed.

All the Comms should have been done via an OFDM-like scheme rather than just a binary sequence over a twisted pair, giving far more throughput and allowing for compensating for conductive dirt in the plug causing crosstalk.

Overall, I believe such a design would have reduced costs, since there is no longer a need for such precision on cable and plug manufacture, which more than compensates for a tiny OFDM 'modem' inside the USB phy.

TIL, one more terrible design I can thank Google for!

Maybe in another 30 years we’ll get a properly symmetrical design with something approaching lighting’s durability.

The properly symmetrical design already exists and it is the phone connector, e.g. TRS

Rough consensus, running chargers at work. [1]

[1] https://en.m.wikipedia.org/wiki/Rough_consensus

Reply on news.ycombinator.com