You know what's nice about USB3 ports though? They're also USB1 and USB2 ports, so they work with like every peripheral made for the last 20 years.

Honestly, I have a new Dell XPS, and in my opinion it has the ideal port combination: 2x USB3, 1x HDMI, 1x USB-C. I'm covered for the past, present, and immediate future. The only loser is a handful of peripherals that plugged into Thunderbolt on my old MBP (basically an Ethernet and Firewire adapter).

You know what's nice about Type-C ports though? They're also USB1 and USB2, and USB3 ports (with a cheap adapter), so they work with like every peripheral made for the last 20 years.

- my cassette player is a CD player (with a cheap adapter)

- my VGA port is an HDMI port (with a cheap adapter)

- my wall outlet is a micro-USB port (with a cheap adapter)

- my Lightning port is a headphone jack (with a cheap adapter)

Adapter-based compatibility is so convenient.

All of these are active adapters USB-C to USB-A is passive(except a resistor). You can use USB-C to MicroUSB-B instead of USB-A to MicroUSB-B cables if your device is not hardwired.

> Type-C ports

> They're also USB1 and USB2, and USB3

> work with like every peripheral made for the last 20 years

too bad there is no guarantee they will work with devices with Type-C ports as the type-C port is a mess and there is no mention on the standard on how to communicate capabilities of the port. Can it do thunderbolt? who knows?! can it charge the device? who knows?! Can it run PCIe lanes? who knows?!

Actually the basic USB fallbacks and simple charging are all handled through passives (resistors) and 100% of Type C cables will support them. Sadly that's the only thing you can depend on by inspection.

Communication of port capabilities are described in the spec; not sure about cable capabilities though.

You'd think that companies couldn't mess up a simple resistor. https://docs.google.com/spreadsheets/d/1vnpEXfo2HCGADdd9G2x9... proves that wrong.

The charging thing in particular seems like a disaster— the situation is bad enough with dedicated power supplies, where we have Amazon pages listing specific devices and laptop models that are known to work; but then there's this whole classes of setups where the correct behaviour isn't even knowable without directly querying the user. If I plug that USB-C battery bank into my laptop, should it be charging the bank from the laptop, or charging the laptop from the bank? Does it depend on whether the laptop is plugged in or not? If the laptop is plugged in to a current-limited source, should it prioritize charging the bank or its internal battery? Gaah.

Here's what the Android prompt for this looks like: https://www.quora.com/Devices-can-charge-or-be-charged-via-U...

Aren’t the adaptors for USBC quite a bit more expensive? Is t there essentially a chip in the cable?

No, adapters from USB-C to the older USB ports are completely passive, they only have a pair of resistors to identify as an adapter.

To be fair, you've certainly taken your time; USB 3.0 has been with us for about 8 years at this point.

Out of the 4 computers I use regularly, only my work laptop has USB 3.0 and it was given to me less than a year ago.

Taking my time? The computers work exceptionally well. USB 3.0 doesn't seem reason enough to replace them, and much less does USB C which is not compatible with any of my devices.

USB-C is a revelation - power, display, storage, all through one cable.

Bah, the big innovation is the reversible connector. Not sure why that took so long.

Exactly. The old USB connector is just horrible... not only is it non-reversible, but it doesn’t “feel” good. I swear my average number of attempts to plug one in is >2, because even if you get the right orientation, if the alignment isn’t perfect it feels like it’s wrong.

The Type-C connector isn't perfect either, if it is only slightly misaligned my screen goes black, sometimes a small bump to the connector is enough. Good thing I don't have external drives connected to it (yet).

Shrinking these connectors has gone too far.

If you look back through old stack overflow questions about the reversibility you’ll find people saying it can’t be done any other way, giving all sorts of demonstrably false arguments about it being impossible to arrange wires like that!

I never understood why something like 4-pole 2.5mm phono wasn't used for USB. Then you'd have no orientation issues at all, and fumble-insertion would be trivial.

An argument I've heard against this is that a rotating connector wouldn't be electrically stable enough. But this is easily addressed - you could easily add a toothed collar to the plug preventing it from rotating, at the small cost of there being a finite number of dozens of possible orientations, instead a continuous infinity. Or better, just design the protocol to be tolerant to packet loss.

Audio jack has this problem that it shorts contacts during insertion/removal. It wouldn't be compliant with EMI too.

Even if they didn't want to make it reversible they could have made it...something like a rectangle. VGA cables are easy to see the orientation of since they are trapezoidal.

I haven't seen these arguments, but not passively reversible perhaps? (Isn't there logic in the cable or something like that?)

Just picture a USB-A connector. You know the metal plates on the tongue? Put metal plates on the underside as well. Connect each plate on one side, to the plate on the other side, but at the reverse end of the tongue. There you go - passive reversible.

Passive reversible USB-A is actually rather common, as a quick search will show; I first noticed this on USB drives (double-sided tongue), but cables are available with those connectors too. The reversible females look quite a bit more fragile than the males; the latter is simply a double-sided tongue, while the former has the contact block thinned to a double-contacted septum.

Reversible micro-B is also a thing, I recently bought such cable.

I read in an interview with the original USB instigator that he wanted reversible cables but in order to get a standard organized and agreed to they had to go for a non-reversible cable to cut BOM cost.

That's a revelation? Thunderbolt has had all that for nearly a decade now.

But that's not the point, is it? The point is that after 8 years, it's exceedingly reasonable for a standard to be superseded.

No, it's not. The longer a standard lasts the better. 8 years is really short. I can still connect new earphones to the jack of my CD player from 20 years ago, or my even older mini-cassette player. That's a good standard. What kind of standard lasts less than 10 years?

I mean, if you want to stick to the 1.5 MB/s transfer rates of USB 1, more power to you.

Analog audio technology was basically perfected 40 years ago with the Walkman. You couldn't develop a significantly better connector today, even the size was optimal for small portable devices.

Data transfer technology OTOH is still improving by leaps and bounds, and the old ports are too big for modern portable devices like tablets and phones.

Agreed! Now XLR, THATS a good standard. Been around since the 1950s at least, and still the de facto connector today. Sure, you need to insulate against background interference and account for ground hum, but honestly that makes me for a clean connection when done right. XLR and Ethernet is all we need.

i still pine for the world envisioned by the first PoE working group; a world where every house's power outlets are just rj45s.

Wow! Never knew this was a thing. Would that be enough to power heavier load appliances (toaster ovens, window AC, etc)?

it's at that state now, yes. 100W PoE is standardized and available on the market. the standards body dragged their feet for a couple decades, vendors implemented their own proprietary high-power PoE injectors and devices, and a lot of the momentum behind PoE died out due to incompatibility hell. imagine every power outlet in your house was networked, could support gigabit data transfer, and the whole thing was so safe you could jam a fork into the outlet whenever the whim struck you -- that was the original dream.

60Hz Electric Grid: 122 years in service.

NEMA 1-15: 115 years in service.

IPv4: 41 years in service.

IPv6: 24 years in service.

Seatbelts: 53 years in service.

Some things are better not replaced every 8 years.

I super agree with everything, with the probable exception of Seatbelts. Is that even a standard with a concern for compatibility? I'm not sure if I can plug the belt of one car to the seat of another. Would be cool, I guess. If my seatbelt ever breaks, I guess that would mean I can obtain a "standard" seatbelt without worrying about compatibility.

There are "heavy" people who carry a seat belt extender because most cars don't have a long enough seat belt without it. So yes there is a compatibility concern.

Though there isn't a standard. Different manufactures use different connectors, not to mention air planes.

There's a standard in that cars must have them, and they must have certain properties in terms of abrasion, strength, surface area etc.

I agree with your general sentiment but NEMA 1-15 must go (it kills a few people each year) and so must IPv4 (it spawned NAT which destroyed the Internet).

My Logitech wireless headphones don't work well with USB 3.x, they only output stereo sound.

It is required to use an USB 2.0 port to get surround sound.

USB 3.x is also unreliable to no end. My external disks, all using USB 3, only worked with the very fast speed for a week or so. Then the cable dies and it is basically just as fast as USB 2.

So, 8 years of unreliability. What a terrible record.

My desktop is plenty powerful, but doesn't have USB 3. Nice to know I can skip it and just get USB 4 on my next mobo.