Hacker News new | past | comments | ask | show | jobs | submit login
How many kinds of USB-C to USB-C cables are there? (kernel.org)
407 points by zdw on July 15, 2019 | hide | past | favorite | 204 comments

After buying a screen which has only a USB-C DP-Alt-Mode input, I was recently in the market for a matching cable. My only requirements were “it should be as long as possible, at least 2 meters” and “it needs to support DP-Alt-Mode”.

Now I thought this would be easy, just ask Google what kind of cable I need, right...? Well... no. Not at all. I wasn’t able to find any information whatsoever what kind of USB-C cable is needed for supporting Alt-DP-Mode. I guessed (really guessed, couldn’t find information) a USB 2.0 USB-C cable wouldn’t be good enough. But is a 5Gbps cable enough? Or is a 10Gbps cable needed? Or maybe neither and I need to buy one of those really expensive USB-C Thunderbolt 3 cables...?

In the end I just bought a 2 meter USB-C 3.2 5 Gbps and just hoped for the best. And indeed it works. Yay!

The thing is... I’m a technical person, and even I can’t figure this out. How is a non-technical person supposed to navigate this?

And by the way: one piece of information I did find about USB-C and Alt-DP-Mode is that the current spec for Alt-DP-Mode doesn’t allow use of active cables, so length is limited to 2 meters. That’s another pitfall restriction.

Why don't they give NAMES to these different types of cables?

They seem to have either no names or too many names where it's unclear what is intended for the consumer.

Just giving the cable a clear name (in words, not icons), so you could know what you were buying, and a device could say "you need this kind of cable", and a person could write a blog saying "to support DP-Alt-Mode you need X kind of cable" and all use the terminology... would seem to go a long way.

It seems odd that the USB people have not figured this out.

We can't even figure out what the CALL em all consistently.

> Why don't they give NAMES to these different types of cables?

Exactly! We can call the one for small peripherals "USB", the one for monitors "VGA", the one for networking "ethernet", the one for power "AC"... I suspect most users can figure out how to manage that.

Snark aside, I actually agree with you. The market powers that be completely messed this up. Basically, phones have limited port space. And for their benefit, and only their benefit, we've tried to cram together the entirety of device I/O into one multi-pronged standard that was clearly never going to work.

And those phones? Their fancy USB-C ports are almost exclusively idle. A median high end phone never does anything with that port it couldn't have gotten with a single barrel connector with ~10V DC.

Camera connectors and card readers work on most Android phones, and it's awesome to be able to swap them back and forth with my laptop. If you work at it a little so do USB drives and yubikeys. I can also charge my laptop, phone, camera, headphones, and tablet from one charger.

I'll take 10 minutes extra effort every few years to order the correct cable in exchange for that...

> Exactly! We can call the one for small peripherals "USB", the one for monitors "VGA", the one for networking "ethernet", the one for power "AC"... I suspect most users can figure out how to manage that.

Wrong kind of names. For these cables, all you should need to know is a speed rating. And they almost work that way, but the ball got dropped in a few places and the cables are badly labeled, so it turns into a nightmare.

> And for their benefit, and only their benefit, we've tried to cram together the entirety of device I/O into one multi-pronged standard that was clearly never going to work.

I can't agree at all with that. The needs of phones were already satisfied by USB micro, long before any of this cramming started. USB-C's complication is not because of phones.

And even though the data connection is rarely used, it sucks to not have it available for file transfers and backups and tethering. And barrel plugs are terrible for charging on the go. You have far fewer outlets than USB, and there are too many different sizes.

Also, barrel jacks don't have just one polarity. I learned that the hard way when I fried my guitar amp as a kid.

They usually have the polarity written down, in a tiny font on a label back down on the charger. That's why I always put labels on barrel connectors, since the manufacturers can't be bothered.

Didn't EU legislation have some impact on this?

Legislation set the USB-A connector on the transformer, but not the other end of the cable.

Before the legislation, most phone chargers had the wire fixed to the transformer, and were unique to each manufacturer. The law reduced the e-waste from obsolete chargers, but didn't care so much for the wire.

This doesn't match what German Wikipedia says:

Der Micro-USB-Standard bezeichnet eine Regelung auf Basis des USB-Standards, nach der ab 2011 die Steckverbinder für die Stromversorgung von Mobiltelefonen und anderen elektronischen Geräten in Europa vereinheitlicht wurden (Norm EN 62684:2011). [...] Das Kabel kann entweder fest oder durch einen Stecker mit dem Ladegerät verbunden sein. Bei einem steckbaren Kabel ist ein Standard-USB-Stecker (Typ A) zur Verbindung mit dem Ladegerät zu nutzen.

The Micro-USB standard is a regulation on the basis of the USB standard, according to which the plug for charging mobile phones and other electronic devices has been unified since 2011 (European Norm EN 62684:2011). [...] The cable can either be fixed to the charger or connected with a plug. In the latter case a USB-A plug is to be used.

"... An Adaptor can also be a detachable cable." and Apple's non-standard connector being compliant suggests this is either permitted, or a very wide loophole in the legislation.


> Why don't they give NAMES to these different types of cables?

The whole catchy idea behind USB-C (unfortunately perpetuated here by people who should know better) is that there is ONE SOCKET, ONE PLUG and ONE CABLE. Oh, the dream of using a single connector.

Unfortunately, this is far from reality and the result is a complete mess, which is why I'm not a big fan of the whole USB-C thing. Especially given that I'm a hostage, because a certain computer company whose computers generally suck the least, decided to show courage by moving to USB-C even though the ecosystem is nowhere near mature — let's put pressure on the manufacturers by getting hostages (oops, "users") on board.

I wish USB-C mandated some sort of connector marking: color strips, or whatever, that would let me figure out what would work.

I wish USB-C mandated one cable. Just one. If you cut anything out of it, if you did anything out of spec, it's not a valid USB-C cable anymore. You either have all features, or no features.

There's nothing benefiting customers from this cable mess. It only enables dishonest businesses, as customers can't tell cables apart until they plug them in and try to use them.

You would get usb-C bootleg cables with missing features. Instead of having USB-C branding on them, they'll have generic 'works with any modern phone!' etc.

I mean cable with all the features is damn expensive if you compare it to the usual(shitty quality) USB2.0 cables. Most folks would buy cheaper one because difference in prices between USB2.0 and USB-C with all features would be too high. It would feel like a rip-off to them.

You'd still get the problem of devices either supporting various modes or features or not.

The whole idea of "one connector that does everything" is flawed, unless you standardize what "everything" means and give users visual clues. Such as, oh, a different type of connector, like we used to have.

I get precisely ZERO benefit from using a USB-C connector for my monitor, I only get a large amount of pain.

You're right.

The line I'm trying to stride here is, on the one hand, avoiding what USB-C ended up being, but on the other hand not returning back to the days of feature phones, where you had as many physically and electrically distinct charger and data ports as there were phone models. There must be a sane state somewhere in the middle.

There is. USB 3.0 micro.

That would certainly mean cutting features. Thunderbolt limits the length of the cable to ~50cm, but that's way too short for a power cord.

Thunderbolt 3 cables can definitely be different from fully featured USB-C cables though. They're also labelled differently, with the Thunderbolt lightning bolt mark.

I read somewhere that they are going to merge Thunderbolt in the USB 4 spec, so they won't be so different soon :)

> Unfortunately, this is far from reality and the result is a complete mess, which is why I'm not a big fan of the whole USB-C thing. Especially given that I'm a hostage, because a certain computer company whose computers generally suck the least, decided to show courage by moving to USB-C even though the ecosystem is nowhere near mature — let's put pressure on the manufacturers by getting hostages (oops, "users") on board.

I still remember when the iMac came out with two USB ports, a CD drive, and nothing else. It was a couple of years too early, or did it prove that you could make a successful computer without vital things like a parallel port and a floppy drive?

> Especially given that I'm a hostage, because a certain computer company whose computers generally suck the least, decided to show courage by moving to USB-C

Since when was USB-C such a major issue on Thinkpads?

They did, kind of. USB2, USB3.1 gen 1, and USB 3.1 gen 2 are 480mbps, 5gbps/DP-alt-mode and 10gbps/DP-alt-mode, respectively. The problem is really that people won't remember that distinction unless they need to know it regularly, and unless you know to Google "difference between USB 3.1 gen 1 gen 2" you won't be able to figure it out.

All they need to do now is to drop the name USB in favour of SPC/OSB - Single Purpose Cable w/ Optional Serial Bus.

The whole thing makes a mockery of the core reason for USB to exist, which was so we wouldn't need a unique cable for each device. And so here we are...

Thats absolutely right as the U stands for universal, and theres nothing universal in it anymore

Have you ever seen those names on a box or Amazon listing?

I have been told those names were not intended for the consumer. (And indeed "USB 3.1 gen 2" vs "USB3.1 gen 1", seriously?).

I think there is another set of names, also inconsistent and unparallel with each other or the first set of names, intended for the consumer, that nobody can seem to figure out.

More precisely: "Why don't they give them _reasonable_ usable names, and only one set so everyone uses the same one?"

This still seems to be an obvious part of the design of the product to me.

I think the GP also meant more memorable names.

Would be easier to rename "USB 3.1 gen 2" -> "USB 3.2".

And the latter two would be "USB 3.3", "USB 3.4", problem solved :)

They really only ever make it worse.

Initially, there was USB 3.0 at 5Gbps, USB 2.0 at 480Mbps, and USB 1.1 at 12 Mbps.

Then, when they added a 10Gbps mode, they called it USB 3.1 Gen 2, and renamed USB 3.0 to USB 3.1 Gen 1.

They already have plans for USB 3.2: 5Gbps will be USB 3.2 Gen 1, 10Gbps will be USB 3.2 Gen 2, and 20 Gbps will be USB 3.2 Gen 2x2. Yup, Gen 2x2 is next.

And there is no more USB 3.0 - presumably no vendor wants to have to sell an older number, so they rename them all so consumers are confused and think they are all the latest 3.x

ref: https://arstechnica.com/gadgets/2019/02/usb-3-2-is-going-to-...

At this point, I don't blame Apple for holding on to Lightning.

As soon as its phones go USB-C, they're going to have to deal with half a million people on the internet complaining that their junk shop cable doesn't work right or do what they thought it would.

> their junk shop cable doesn't work right

This already happens though. I have a dollar store "Lightning" cable that only has a third of the pins and only works one way.

So apparently, according to TFA, the DP capable cables are "SuperSpeed" and marked with an SS.

Learning just that one fact made it worth the click. I now know to just always look for SuperSpeed (or the SS marking) on any cable I buy.

If we could just all start calling them SuperSpeed cables, SuperSpeed10 for 10Gbps, and I assume we will eventually have SuperSpeed20, etc... that would be just fine by me! Honestly, better marketing IMO as well.

The time for the "USB" acronym has passed. Especially since the debacle of "USB3" and Gen1 Gen2 nonsense. "SuperSpeed", optionally followed by a number which indicates a data rate would be just lovely.

it's all usb cables. that what the usb people thought up and thats how u call em. on that note, a lot of usb cables seem broken :D

The answer is supposed to be easy; it's "you need to buy one of those really expensive USB-C Thunderbolt 3 cables, because those cables—the ones that do everything—are supposed to be the only type of USB-C cables."

USB-C cables are supposed to be universal, i.e. have all the features, all the time; and thus, USB-C cables should (in 2019) be very expensive. Any USB-C cable that isn't very expensive (in 2019) isn't following the standard.

Note that Thunderbolt 3 is not part of the USB 3.2 standard. It's the basis for USB 4, but right now to be fully standards compliant with everything USB-C itself can fully do you do not need to be Thunderbolt 3 compatible, because thunderbolt 3 is not part of the spec at all.

And even if you wanted to be Maximum Compatibility even with Thunderbolt 3 you may not have that option as TB3 tops out at a cable length of 2 meters, and at that length you're dealing with half-rate bandwidth (just 20Gbps instead of the full 40Gpbs - which you only get up to 0.5m)

> Note that Thunderbolt 3 is not part of the USB 3.2 standard.

It may be better to say that Thunderbolt 3 is a 'superset' of USB: TB is its own protocol, which allows for the carrying of USB over it (as well as PCIe and DisplayPort).

> isn't following the standard

This is just plain wrong.

The USB standard explicitly allows cables that have a USB-C connector and don't support any USB features above USB 2.0.

There are plenty of cables that are cheap and support everything that's actually in a USB standard. They only go up to 1 meter, though. Even thunderbolt 3 isn't too expensive over short cables. They don't need to be expensive.

At the other end of the spectrum, there don't seem to be any cables that support both thunderbolt 3 and USB 3 at 2 meters, no matter how much you're willing to spend.

> USB-C cables are supposed to be universal

I may be wrong here, but from what I understand this doesn't seem to be the case (DP, Thunderbolt, etc. are all "optional"). It would also mean that all cables would be limited to 50cm.

There was the design intent, and then there was the final standard.

The USB-IF is composed of manufacturers that wanted to be able to cheap out on cables; so of course they made all the expensive features optional, at the expense of anything USB-C was designed to accomplish.

In an alternate world where USB-IF was headed by a benevolent-dictator-for-life, USB-C wouldn't have any optional features. (It'd also be in nearly nothing, because no manufacturer would want to pay the BOM costs of supporting it.)

> (It'd also be in nearly nothing, because no manufacturer would want to pay the BOM costs of supporting it.)

It'd be everywhere if manufacturers had no other choice. Mass production would keep the price low.

In that reality, Apple would be actually courageous for switching to USB-C and forcing everyone's hand. The EU could jump in to, and tell phone manufacturers that if they want to switch away from plain old USB, they can switch over to USB-C, but no going back to proprietary connectors bullshit. This would be a much nicer world indeed.

>Apple would be actually courageous for switching to USB-C and forcing everyone's hand.

Last time Apple was courageous ─ the headphone jack got killed. The word 'courage' was used to onboard Airpods and move the general trend towards sealing off most ingress points on a device; contemporaneously selling an adapter for a pretty penny.

Whilst I agree that some manufacturers followed the above-mentioned trend. They were not pushed, but rather jumped willingly, as it made financial and commercial sense by applying the maxim - where Apple leads, others must usually follow. There were also unsubstantiated rumours of Apple abandoning their role in development of USB-C implementation, perhaps due to the lack of a moat[0]. Nevertheless, I don't expect them to be altruistic in their motives and fully expect them to monetise the USB-C standard via a plethora of cables and dongles, possibly under the umbrella of MFi program, especially targeting the lack of certification and/or the vacuum/ambiguity created by USB-IF, leaving the rest of the field to carry on dealing with the existing issues.

Also, I would much rather put up with the current botched up state of affairs, than have EU involved in meddling with it and make things even worse.

[0] https://9to5mac.com/2015/03/14/apple-invent-usb-type-c/

> Last time Apple was courageous ─ the headphone jack got killed.

Hence my use of the word "actually" - the headphone jack situation is IMO anticonsumer, not courageous.

As for EU involvement, it seems to have worked out for the best with forcing phone manufacturers to use USB for charging - I still remember the days of feature phones, where we were all drowning in phone chargers and (separate) data cables, as each brand and each model generation within the brand used different ports and voltages. It was ridiculous levels of plastic and electronics waste for the sake of business shenanigans.

>As for EU involvement, it seems to have worked out for the best with forcing phone manufacturers to use USB for charging - I still remember the days of feature phones, where we were all drowning in phone chargers and (separate) data cables, as each brand and each model generation within the brand used different ports and voltages. It was ridiculous levels of plastic and electronics waste for the sake of business shenanigans.

You are romanticising the notion of Apple and EU being the standard rod-bearers. Despite using electrical waste and dubious business practices as a reason for EU involvement, it would seem that you are unaware of the saga of the 'common charger' and the villainous role played by Apple.

A key take-away:

The EU's failure to regulate, in essence, provides Apple with an argument to keep their own solution.

For Apple, the common charger is a success story – of regulation avoided.


Can you link to some cables that are certified to implement the entire standard? Even that has been difficult for me to find.

I actually don’t think any of the above commentary is true.

More expensive doesn’t necessarily mean more support ... thunderbolt is not usb-c even tho they have the same connector on the end ...

If you are in the market for an expensive and long thunderbolt cable — caldigit makes one. I just bought it and felt annoyed at paying 50 dollars for a cable — but I’m also glad for the technology needed to make this cable possible to exist so ... maybe I can swallow this as funding the future?


I detected sarcasm. I'm not sure if that's just me or the other respondents so far have mislaid their detectors.

My Samsung phone came with a very thin USB-C cable that only supports USB 2.0 speeds. I think that's following the standard.

«But is a 5Gbps cable enough? Or is a 10Gbps cable needed?»

Here is the answer: if you need DisplayPort 1.0-1.2, a 5 Gbps cable is enough.

If you need DP 1.3-1.4 (where they upped the transfer rate for higher resolution and fps,) you need a 10 Gbps cable.

If you need DP 2.0 (I don't think products are out yet, the standard was finalized last month,) you will need a USB4 20 Gbps cable.

DP1.2a should be enough for 60Hz at UHD/2160p. 1.4 is needed for HDR support, 5K, higher refresh rates, etc.

DP 1.2a goes up to 21.60 Gbit/s, and 4k 60hz (which does work fine over 1.2a) needs 12.54 Gbit/s. I'm not sure how the 5Gbit/s and 10Gbit/s USB-C cables fit into this. It's obviously not just based on the cable's "rated" bandwidth, as if it was neither the 5Gbit/s nor 10Gbit/s connection would work at all for display. But it does, so it must not be, so what is it?

Is it 5Gbit/s per twisted pair? That would make sense as displayport uses 4 pairs so 4 * 5 comfortably gets you the 12.54 with plenty left over for regular data. But then it would matter for the 10gbit/s cables if it's talking about superspeed 1x2 (2x5gbit/s) or 2x1 (1x10gbit/s).

«I'm not sure how the 5Gbit/s and 10Gbit/s USB-C cables fit into this.»

A USB3 cable has 4 differential pairs. When in alt mode, all 4 pairs transmit in the same direction, so a 5 Gbit/s cable can transmit 20 Gbit/s.

I generalize them for friends who ask and break them down into these variants:

* USB 3.1 (Type-C to Type-C) Charge Only * USB 3.1 (Type-C to Type-C) Charge and Data @5Gb * USB 3.1 (Type-C to Type-C) Charge and Data @10Gb * Thunderbolt 3 (Type-C to Type-C) Charge and Data @20Gb * Thunderbolt 3 (Type-C to Type-C) Charge and Data @40Gb

I believe 5/20Gb cables are generally rated at 65W and 10/40Gb cables at 100W. But like you've noted there's no standard labeling. So it's hard to find what you're looking for half the time.

It's just easiest to find Thunderbolt 3 cables since those are backwards compatible and if you buy the 40Gb ones you're going to be safer with regard to plugging both your laptop into a charger and your phone.

I appreciate your response, but holy hell how is anyone supposed to memorize all that to buy a damn laptop or phone charger

Wait a minute, USB-C Thunderbolt 3 cables are also USB-C to USB-C cables, aren't they? Why aren't they included in this article? I guess there's 6 kinds of USB-C to USB-C cables, except for all the other kinds of USB-C to USB-C cables. How many are there really?

TB3 is not part of the USB-C spec, they just happen to use the same connector. The cables are usually backwards compatible.

I know TB3 is not USB-C, but as long as a TB3 cable is USB-C compatible (can you really call it "backward" compatible if TB3 is not USB?), that cable is a USB-C cable. You can use it to connect 2 USB-C devices. That means there is more than 6 kinds of cables you can use to connect USB-C devices to each other, contrary to what the article says.

The mystery might not even be completely solved then, have you confirmed that it's not performing some unwanted compression to get your data over a lower-bandwidth cable?

Maybe a good indication would be to check your CPU usage before and after using 5Gb/s and 10Gb/s cables. Even then it's not a great indication, many modern processors have dedicated modules on the CPU for video compression/decompression (given how large the use case is, it makes sense).

And then you have other factors that may prevent you getting full speed:

* Connection between the cable and the device on both sides (oxidization, scratches, etc)

* Manufacturing flaws or damage during delivery/use in the cable (un-shielded areas, breakages in the cable, imperfections in metals, etc)

* Temperature - if you're pulling 20W through this cable it's not entirely impossible you might start noticing some heat, especially if it wasn't designed to take the power draw (increased heat due to power pull = greater resistance = worse signal)

My pointy is, as we turn up the requirements of these cables, just making sure they deliver on their promises will become more difficult. Most people won't have an oscilloscope that can even measure these cables anymore and just the act of putting the oscilloscope on the lines increases capacitance and changes how the signal looks.

A while back I had issues with two Ethernet cables (CAT 7 I think), both straight out of the packet, one performed at half speed and the other at full. I was tearing my hair out for a while before I realized why the network was bottle necked.

> It needs to support DP-Alt-Mode

For future reference, a Thunderbolt 3 cable (which is USB-C connector) will work.

Not true! Active thunderbolt 3 cables often only support thunderbolt 3.

Per the Thunderbolt FAQ [1]

> What is Thunderbolt 3?

> Computer ports with Thunderbolt 3 provide 40Gbps Thunderbolt – double the speed of the previous generation, USB 3.1 10Gbps, and DisplayPort 1.2.

I'm curious: which active cable did you use that didn't support DP?

[1] https://thunderbolttechnology.net/tech/faq

What you quote is not related at all to what Thunderbolt cables support beside Thunderbolt. It just states that you can use Thunderbolt to transmit DisplayPort signals. By that you're utilizing the Thunderbolt alternate mode to tunnel DisplayPort signals, instead of using the DisplayPort alternate mode [1]. Using Thunderbolt to transmit DisplayPort signals only works with devices which support Thunderbolt, not with screens supporting "just" DisplayPort. To make confusion even worse: All existing Thunderbolt controllers also support the DisplayPort alternate mode, so a lot of people who believe they're using Thunderbolt are probably just using the DisplayPort alternate mode instead. Afaik Thunderbolt cables usually support both, but I'm not sure if that's mandated somehow.

[1] https://en.wikipedia.org/wiki/USB-C#Alternate_Mode_partner_s...

Thanks for the clarification.

What a clusterfuck of a standard.

Note: This cable does not support native DisplayPort 1.2 or Thunderbolt 3 connections to a DisplayPort monitor. https://amazon.com/StarTech-com-Thunderbolt-Cable-40Gbps-Cha...

Regular USB-C monitor is not support, USB 3.1/3.0/2.0 and USB-C is not support https://amazon.com/Thunderbolt-Certified-Nekteck-Compatible-...

USB 2.0 ️[check], USB 3.0 [ ], DisplayPort [ ] https://amazon.com/Belkin-Thunderbolt-Cable-Feet-Meters/dp/B...

Can you point me to any 2-meter thunderbolt@40GHz cable that does support USB 3 and/or DisplayPort? None of the ones in this chart can do it: https://docs.google.com/spreadsheets/d/1vnpEXfo2HCGADdd9G2x9... Only the 1-meter active cable can.

Well shit, I stand corrected. Thanks for the links!

Also the most expensive option and limited in terms of maximum length.

How are the Thunderbolt cables wired? Do they have more than the 10 USB3 wires or is it just a higher standard of cable?

Isn't USB4 = the next Thunderbolt standard?

> After buying a screen which has only a USB-C DP-Alt-Mode input

OT, but what screen did you buy? I'm in the market as well for a USB-C powered travel monitor and was about to head down the what-cable-works rabbit hole.

I didn't even know they made 2.0/C cables.

The USB-C Charge Cable that Apple includes with their laptops is a 2.0 cable... (that supports 87W PD)

This only seems to be counting USB-C to USB-C USB 3 cables

The actual situation is much worse when you add in passive and active C-to-C Thunderbolt, MHL, DP, and HDMI cables, some but not all combinations of which may or may not optionally support some but not all combinations of other signals as well — eg) a passive TB3 C-to-C cable will support DP and USB3 (but not MHL or HDMI), but an active TB3 C-to-C cable will be guaranteed to support only TB3, but may optionally support DP.

Visually making a complex matrix of incompatible things look identical has resulted in an absolutely hopeless situation for consumers, who trained by decades of “connector shape == bus” think that USB-C is a monolithic bus that supports all of these different use cases and then don’t understand why they can’t “just”, say, plug their USB-C Nintendo Switch (wired for USB 3.0 w/ a DP alt. mode signal) into a USB-C eGPU (wired for TB3) using whatever “USB-C cable” they have lying around, which could potentially support none of the incompatible protocols either port can handle anyway.

It's a bloody goddamned nightmare. For example, a brief review of my devices:

>Phone A. Type-C port limited to USB 2 data speeds but supports HDMI out via DisplayLink.

>Phone B. Type-C port that supports USB 3.1 speeds but doesn't do video out.

>Tablet A. Type-C port that supports USB 3 speeds but only charges through a MicroUSB port.

>Tablet B. Type-C port that supports USB 3 speeds and charging, but only charges at 0.8 amps through the type-C port and relies on a pogo-pin dock for actual 15-watt charging.

That's not even mentioning all the devices that advertise one thing and then break the standard in a hundred different ways like the Switch.

I don't see your point. Devices are free to support or not support whatever capabilities they want. Having a specific port is no guarantee that the device will use this or that capability offered over that port; mostly it's a guarantee it'll use at least one capability offered by that port.

The problem with USB-C is that many of the cables don't do everything. Nobody else is complaining that the devices don't do everything. Devices have never done everything possible under the communications standards they support.

Would you expect a monitor that supports HDMI2.0, to be a 4K monitor, just because HDMI2.0 supports 4K monitors? I would think not.

> Nobody else is complaining that the devices don't do everything.

Tons of regular consumers think that USB-C is a protocol and frequently complain that things they think should work together don’t, because they’re actually radically different ports electrically that happen to be shaped the same.

Gaming forums are rife with elaborate conspiracy theories behind why Nintendo won’t “just” release an eGPU Dock for the Switch, which it “obviously” “must” support because it’s USB-C and eGPUs have “supported” USB-C for years.

And now, Nintendo has announced a Switch Lite with no TV out (presumably they removed the DisplayPort crossbar IC that does alternate mode muxing with the USB 3.0 signal, to reduce manufacturing costs). But Joe Average Consumer is up in arms, this “must be them sabotaging it in software”, because “video out is a required part of the USB-C standard, this wouldn’t even save them any money, USB-C always physically supports video out”.

The average consumer in no way understands that USB-C is just a connector shape that a bunch of different incompatible things use. They think that, like almost all previous ports, the shape indicates a common protocol that is shared by things that have that shape. And they are absolutely complaining about a lack of what they perceive should be compatibility.

> presumably the removed the DisplayPort crossbar IC that does alternate mode muxing with the USB 3.0 signal, to reduce manufacturing costs

It probably saves some amount of money, but I'd bet that deliberately leaving it out to encourage sales of the full switch is a bigger factor.

Otherwise I agree with you about protocol confusion.

(Would you even need a mux? I don't think the switch uses USB 3 for anything.)

Yeah, you need the mux because you can run a wired Ethernet adapter in over USB at the same time it’s outputing a DisplayPort signal on the port.

That uses the USB 2.0 wires, as far as I can research.

And if it was using a mux to pick signals they wouldn't run at the same time anyway.

Well, I can tell you that it has a mux, anyways — it’s a Pericom PI3USB30532 per iFixit’s teardown. Exactly what for I don’t know.

That's fair.

Though presumably the switch lite doesn't need to support USB at all.

If a device has an HDMI port I know I can use that to connect a display. I don't know details like the supported resolutions, but at least I know that's a supported use case.

If I notice that a device has a USB-C port all I really learned is that it has a USB-C port. Who knows what it does.

But that's most port standards that have ever existed.

What can a device do with a parallel port? A serial port? A SCSI port? A Firewire port? Who knows? It's up to the device. It can certainly send signals over that port, of some kind. But what kind of signals? Up to the device. The host can hopefully probe a device-class out of the client and stand up an appropriate driver to talk to it.

A CD player has a 3.5mm audio jack. Pop quiz: does it support SPDIF? In fact, is it line-level or phono-level output? (An important question to avoid blowing out your equipment!) Answer, in both cases: who knows?

You find a random CAT-6 drop in the wall. Does it contain Ethernet? Does it contain Power-over-Ethernet? Does it contain both? No way to know without asking someone or plugging it in.

Or, for a slightly more interesting example of a "port": what can a random peripheral board do with a motherboard's PCI-e (or PCI, or ISA, or any historical equivalent) port? Well, whatever it likes. Does anyone expect every PCI card to be simultaneously a network card, graphics card, sound card, RAID card, etc., just because those are all things PCI can carry? Of course not. PCI-e is a single port, and a single protocol (sort of), but that doesn't mean it's a single set of exposed capabilities.

There's a difference. In most of those cases, you can expect to pull out any fitting cable for that port and the device will do what it's meant to do. There was a little confusion with 3.5mm jack, but the plugs were trivially inspectable (you could count how many connections they had).

With USB-C, whether or not a device will do all, some, or any at all of its functions, depends not just on the device, but on what you're connecting it to and which one of the countless types of cables you happen to have, all of which look identical to someone who's not an EE with appropriate equipment.

Makes me think, maybe I should start selling USB-C cable testers, because if old USB taught me anything, is that you can't trust cable manufacturers to not lie, even if the cable standard is as simple as it gets.

I feel like there's some confusion over peripheral support, and host support. What can I do with a serial port? Everything. I can talk to every device that ever used a serial connection. Sure, I might need to write a driver, but that's really a separate issue. There's absolutely no problem with "your serial port does not support <gibberish> so you can't use this particular mouse/printer/random industrial equipment".

With your PCIe example, again - I don't expect every card to do everything. But I do expect every card to work on my computer's PCIe bus.

I think one of the things that makes USB-C particularly problematic here is that "plug it in and find out" can be harmful to devices since devices can draw 100W of power through the port.

> since devices can draw 100W of power through the port

Devices can only draw 100W of power, which is 5A at 20V, after negotiation; before negotiation (or if negotiation fails), only 5V is available, like in the older USB connector types.

Well that's the spec but I've personally had a Google Pixel melt its own charging port when plugged in with a cheap USBC cable off amazon.

> But that's most port standards that have ever existed.

That's also the problem that USB solved. Until USB-C came around.

My monitor has HDMI ports, but doesn't run full-resolution through it, because HDMI doesn't support that throughput.

So I'm not sure having HDMI is better or worse. I would prefer having slightly cheaper monitor, if anyway I'm going to buy proper DisplayPort cable.

I'd be complaining like fuck if I had a mobile device with a USB-C port that can't be used for charging. That's the most ridiculous thing I've heard all day.

> Devices have never done everything possible under the communications standards they support.

Yes they do - that's the point of a communications standard. Any device with a USB 3.0 port will work with any USB peripheral ever, given suitable drivers. Your HDMI example makes no sense because it's backwards; of course I don't expect an HDMI2.0 monitor to be 4K, any more than I expect a flash drive to act as a keyboard - but I absolutely expect a port that supports flash drives to support keyboards, and I expect a computer with an HDMI2.0 port to support 4K monitors.

The departure with USB-C is that it isn't a communications standard. Unlike any previous USB port, having the port is not a guarantee that it will work with any peripheral with that plug. The situation with the cables merely reflects this fundamental usability flaw - even if all cables did everything, assessing compatibility would still be confusing.

> Having a specific port is no guarantee that the device will use this or that capability offered over that port

And that's the serious underlying design flaw.

Okay so you've a bunch of different devices that have different capabilities - can you also mention the different cables you had to get (I'm just curious).

In my case I have a switch, 2 macbook pros (2018, 2019) and an iPad pro (11.5") and I've not had much of a hard time finding cables, at least for power.

The only cable issue I've run into is that there are cheap MicroUSB to type-C adapters that are sold without a necessary 56Kohm resistor, which limits its functionality. It won't charge at high speed and it also prevents type-C devices from using OTG mode to talk to accessories like flash drives, mice, and cameras. There are also a few quirky combinations, like the fact a OnePlus Dash charger will absolutely refuse to provide power to a Switch while most phone chargers will power the handheld portion just fine if not the dock. A little too well in fact, as I had a RAVPower "40W 3A" car charger melt while using it to power a running Switch in the car.

Nintendo mever advertises Switch as USB-C in any of their marketing though

I think the overall theme of this thread is that it used to be, if the cable fit in the hole it will probably work.

USB-C is the first time most people have run into their cable's shape not corresponding to the cable's capability

(folks trying to do 4k video with shitty HDMI cables excepted)

The Apple Thunderbolt 3 to Thunderbolt 2 adaptor supports nothing except Target Disk Mode.

I tried using it to connect a USB-C SSD to my MacBook Pro 2014, using a Thundebolt 2 M-M wire, and the adaptor plugged into the SSD enclosure. It didn't work, though thankfully it didn't fry anything either.

The Thunderbolt adaptor also doesn't work with MiniDP-VGA or MiniDP-HDMI adaptors. So I would need to buy a whole new set of video adaptors if I upgrade my laptop. Needless to say, I'm not in a hurry to do so.

I like adaptors, and I have a large collection in my bag already for video out/in, phone charging, FireWire/Thunderbolt disk mode, and more. But I haven't even started collecting USB-C adaptors because it would only be for friends ("Sorry for the technical difficulties with the projector...") and I don't know which ones will work.

> The Apple Thunderbolt 3 to Thunderbolt 2 adaptor supports nothing except Target Disk Mode.

This isn’t correct. I use it daily to connect a MacBook Pro with Thunderbolt 2 ports to a RAID enclosure/dock with Thunderbolt 3 ports via a TB2 cable terminated with this adapter. An external monitor is attached to the enclosure, as well as several USB devices. The display port alternate mode signal and USB hub functionality all works simultaneously with the RAID.

Of course, half the problem with USB-C is that the failures are incredibly opaque, so it’s hard to say why the adapter failed in your case. Maybe you plugged it into a USB 3 USB-C port instead of a TB3 USB-C port? My enclosure has separate USB-C ports for each protocol, and I need to be careful to select the right one.

I use the TB3-TB2 adaptor with their old TB2 Gigabit Ethernet adaptor all the time, so that works at least.

I've had a USB-C capable device (MBP) for a few years now, and it originally was extremely annoying. Needing dongles for _everything_ was annoying, and getting rid of Magsafe meant a lot of my existing chargers that I'd accumulated no longer worked. My work laptop (at the time) didn't use USB-C, so I had to keep two chargers around, and I couldn't charge my iPhone from my laptop at all.

Fast forward a few years and I think it's great.

I have a dock at home that I can plug both my work and personal laptop in and have it work seamlessly with two 4k monitors, gigabit ethernet, mouse, keyboard, speakers, etc.

After biting the bullet and getting some USB C -> Lightning cables, I actually need to carry fewer chargers, fewer cables and everything sort of just works. The new iPad Pro helps a lot with that as well. USB PD means that everything charges faster, as well.

USB-C is sort of a nightmare until it's not, and that transition is basically an entire product cycle (or two) for people, which means a lot of time. At some point, I hope that issues like this become more identified in software (as mentioned), and cheap USB C cables that only have 2.0 support are less common.

Furthermore, while there was once a point where the physical layer matched up well enough with product support, actually connecting a display has always been a pain in the ass. Devices are still shipping with VGA ports because of it - USB-C is no panacea, but I think it's a huge improvement in aggregate.

With a MacBook Pro it’s not so bad, because rather than try to save a few cents Apple wired every USB-C port for everything that it supports, so there’s no real distinction between ports (but you still need to distinguish between very similar looking cables that handle, say USB 3.0 in one case and an active TB3 40 Gbps in another)

It’s at the low end of the market that USB-C is an entire mess. This port on your laptop can connect to the charger, but this one doesn’t. All of them do 3.0, but this one does 3.1. This one does DisplayPort but not that one. None do Thunderbolt 3 but you had to read the specs twice and guess that from the fact it’s just not mentioned. Hard for regular consumers to understand, and remember, when it all looks the same.

Except that even Apple forgot certain options in their 50$ USB-C charger and it is therefore non-compliant with specification. Apparently USB-C is hard.

source: https://twitter.com/USBCGuy

Yeah that’s a whole extra layer of fun. On top of the incompatible matrix of protocols you’ve got all of these devices (the Apple charger, the Nintendo Switch Dock, the Raspberry Pi 4, probably any number of cheap gadgets with bad USB-C ports) that don’t even work with things they should be compatible with.

USB-C is the wild godammned west.

On some Macbooks, the USB-C on the left and right side don't support the same max bandwidth

If I'm remembering correctly, it's not that far off from the (blue?) color coding with the USB A connector to indicate USB 3 support for particular ports? Either way, when power gets involved as well, that is a terrible user experience.

I had a friend bring over a Dell laptop with a USB C port that I naively assumed had TB3 support (it was listed as being "optional" on Dell's website). Turns out that the only way to actually know was to either watch it not work with a Dock, or to go into the BIOS and witness some missing settings that indicated TB3 support was not present.

For better or for worse, though, that's a manufacturer problem - they could have easily just not allowed _any_ USB C charging and stuck with an AC adapter, or spent the extra coin to make all of the ports work with a TB controller, etc. Companies have been making user-hostile decisions in the name of cost cuttings (1366x768 screens in 2018) forever, it seems like USB C is just another avenue for that.

My understanding of Dell's implementation is that the TB port has a little thunderbolt next to it while the usb-c one doesn't.

Not to be confused with the little thunderbolt next to some USB ports that indicates charging capability.

Clearly, that's a lightning bolt rather than a thunderbolt. Very very frightening.

Not to be confused with Apple's Lightning connector, of course...

> For better or for worse, though, that's a manufacturer problem - they could have easily just not allowed _any_ USB C charging and stuck with an AC adapter

For what it's worth, I personally would be thrilled by such a decision. I'm against having my power supplied through a data cable and data port. In theory it's safe, of course, but in practice people have fried equipment with bad cables. It's a hard standard to get right. It's as if we were all using Rust for systems programming up until 2016, when we all switched to C because it was "more powerful". In theory you can write safe C. In practice ...

I also think display connections should have their own interface. It needlessly complicates the protocol when we already have perfectly good connectors for this, HDMI, micro-HDMI, Displayport...

USB should be a dumb protocol with a passive cable. The only distinguishing feature between different USB C cables should be transfer speed, and different rates should be given a clearly labeled version number and distinguishing color, like USB 3.1, 3.2 etc.

Of course no one's actually going to implement this, probably the companies involved profit off of incompatibilities. Whenever you have to buy a new cable or new device because it's not supported by the old standard, it's proof this approach is winning.

Paying for extra stuff and more expensive cables, for what benefit again? I could throw out all my USB peripherals and buy new ones that use USB type C, but at the end what have I achieved aside from setting a few $100 bills on fire?

For me, it's the difference between having to plug in at least 5 things when I set my laptop down at my desk (or use a very specific dock that only works with a single size and model of laptop) to just plugging USB C in and having my laptop charge and function with every device. This is worth the price of admission, though the decision was made for me by Apple.

Beyond that, the additional wattage that USB C PD provides means that I can charge my iPhone and iPad significantly faster than even with Apple's "fast" 12W charger previously. This is really nice when I am traveling or just forgot to charge my iPhone.

In addition, now that more devices are coming with USB C support, I can charge everything from my laptop to my Bose headphones with the same cable and charger, which would have not been possible before.

> I have a dock at home that I can plug both my work and personal laptop in and have it work seamlessly with two 4k monitors, gigabit ethernet, mouse, keyboard, speakers, etc.

Umm, so do I, but USB-C isn't involved.

Out of curiosity, what USB-C dock do you have? I am in the market for one.

I (and a number of people I know) have a CalDigit TS3+ and it's worked extremely well for me: https://www.caldigit.com/ts3-plus

I can't speak to what else is on the market, but there's not more I'd want from this.

Can second the recommendation. When I was looking, I think it was the only USB-C/TB3 dock that supported multiple 4k displays at 60hz (one via Displayport and one via a USB-C to Displayport adapter). Their website looked a bit fly-by-night, but I've been happy with the product.

The only issue I've had is that my Macbook sometimes flips my two displays, but I believe that's an OS problem.

Damn, it looks perfect except that it can't do 4k @ 144hz (but to be fair I haven't found a hub that can, only one-off USB-C to DP connectors).

I have the OWC Thunderbolt 3 Dock and it also works perfectly.


Found this[0] some while ago on reddit. It shows the mess that is the USB-C cable market in a spreadsheet with every feature each cable supports.

It’s crazy.


So frustrating.

One would think that after nearly 2 decades of consumer confusion over USB cables that it would be possible to just mark the cables in a clear way with a non-confusing, easily looked up logo (or better, with a alphanumeric-code).

It makes me wonder if manufacturers are just doing this to increase the churn of cable sales, so consumers end up buying more cables because "some work and some don't".

That or it's a spectacular example of committee design failure.

I can't tell what runs deeper: evil or stupidity. :D

"Sufficiently advanced incompetence is indistinguishable from malice."

I don't have an attribution, but it's an apt quote.

> I don't have an attribution

This is called "Hanlon's razor" and, by the way, it is probably a silliest, baseless, most overused saying on the Internet...

No, the original razor is “never assume malice before having ruled out ignorance and stupidity”. This is an ironic (and funny!) riff on it.

I don't think it's malice or stupidity, but rather a desire to fix people's problems and get it right this time. xkcd sums it up best: https://xkcd.com/927/

Even better, why not just make incompatible cables impossible to intermix by, say, using different ports?

Well, cables in question are mostly backwards-compatible, so as long as it would actually transition forward, the prospect of using a single ultimative cable with any device with identical ports seems actually nice. The problem is not that USB-C is a bad standard trying to combine too much in a single port, the problem is that there are multiple different cables absolutely identical from the outside at one time on the market and not only they are not color-coded (as it was with USB3 long time ago) to be distinguishable, manufacturers don't even bother to write on the box what exactly standard/generation this is. That is, even when you specifically buy a cable in a box, not even to mention situations when some piece of wire came to you with a phone or whatever.

It only seems like a nice idea if you don't care about efficiency. If your tool can do all the things then it's going to be a very expensive tool (or will suck) compared to something that's specialized.

I might believe you that USB (or Thunderbolt) is imperfect both to transmit video to my display and to charge my phone, but the thing is, it does it well enough, and even if using some special charger will give me twice the speed, I honestly don't care, the life quality that simplicity of using 1 cable gives me far outweighs the hypothetical efficiency advantage of 10 specialized cables. The only case when I can consider a dedicated output on a consumer-grade hardware is when the signal is fundamentally analog. Otherwise, why should I even care, if both the bandwidth and latency are mostly fine and voltage is ok to power up my device?

But it's not one cable, unless you pay $30-$50 for that single cable. That cable will also be very limited in length. If you want something that is longer or cheaper then you're going to have to get multiple cables.

I wish they would have also discussed how this relates to Thunderbolt 3, which I believe would add at least two cable types: "7 (TB3 passive)" and "8 (TB3 active)"... of course I'm probably completely misunderstanding this crazy labyrinth of lookalikes.

Here's a lovely statement regarding TB3 from an online FAQ:

Are all Thunderbolt 3 (USB-C) cables created equal? No, in fact, there are two types of Thunderbolt 3 (USB-C) cables, passive and active. Passive cables cost less and can reach up to 40 Gb/s data transfer if the length is 0.5m or less and 20 Gb/s if over 0.5m. Active cables can reach up to 40 Gb/s data transfer with a length of up to 2m.

Thunderbolt gets to be an entire extra dimension, thanks to cables that support thunderbolt but not USB 3.

I did NOT know about the Tree, SS Tree, and SS Tree 10 logos actually meaning something, but if that's what's required roughly a quarter of the problem solved.

Another quarter would be mandating clear markings for the Power Delivery capability: I suggest PD{amps}A OR PD{watts}W as text markings, and also the use of some color code; yes both.

The other 50% of this would be to REQUIRE that marketing use specific terms so that search engine results would be useful and to BAN re-use of those same terms if a feature is not supported.

Ideally that last step would include a suffix 'max' for indicating the highest supported tech protocol.

E.G. the current cable could claim (some lesser modes skipped) USB SuperSpeed, USB SuperSpeed 10 max, PD3A, PD5A max, PD15W, PD60W, PD100W max

While a lesser cable might drop bits from the enumeration of the specification. It would also allow searching by a wattage/amperage (in case it matters: E.G. some devices might support 5A but not the higher volts, others the reverse).

There's always pain after a USB spec gets released. This has been pretty much par since USB 1.0 - switch chips were flaky back then, USB drives were quirky and getting them working on anything other than a specific version of Windows was a crapshoot, etc.

Eventually all-in-one chips get less buggy, cable manufacturing gets cheaper and we move on to the next spec level and repeat.

Yes, and I'm looking forward to USB-C ubiquity in the near future, but these are definitely signs that this rollout could have used some more thought put into it. I'm for "USB-C all the things" but I have sympathy for the gripes about knowing what to use when.

I'd prefer if articles like this tried to nudge the tech community towards using the (imho) clearer nomenclature recommended by USB-IF: “High Speed USB” for 480Mbps products, “SuperSpeed USB” for 5Gbps products, “SuperSpeed USB 10Gbps” for 10Gbps products and “SuperSpeed USB 20Gbps” for 20Gbps products, instead of USB X.Y Gen Z which is sort of ambiguous.

This part at the end was intriguing...

> Cables 2-6 are required by the USB specification to include an electronic marker chip which contains vital information about the cable. The host should be able to read that eMarker, and identify what its data and power capabilities are. If the host sees that the user is attempting to use DisplayPort Alternate Mode with the wrong cable, rather than a silent failure (ie, the external display doesn't light up), the OS should tell the user via a notification they may be using the wrong cable, and educate the user about cables with the right logo.

But how is the host OS going to notify me that the cable is the wrong kind of cable to support display output, if it doesn't even know that I'm trying to connect a monitor on the other side? Can I fallback to legacy USB2 speed to identify the peripheral and then decide a higher speed cable is required?Would it have to be based on the user action of trying to toggle the primary display (i.e. Windows-P), triggering a prompt assuming a monitor was connected and that the cable is the wrong type?

Perhaps rather than the host trying to guess that, it would be better for the peripheral to detect on its side that the cable is the wrong type, and then displaying a very helpful full-screen slide explaining the issue. In the case of a monitor, it could even include a QR code to scan with your phone to bring up further documentation, complete with Amazon affiliate links to buy a better cable, you know, because you know, monetize all the things. </s>

I recently bought a USB-C extension cable (so that I could have a USB-C port nearer at the front of a desktop computer instead of the back) and it came with this label stuck to it:

> USB 3.1/3.0 is supported on both side. USB 2.0 is only supported on one side. If it doesn't work, please reverse it.

"Reverse" here meaning unplug one end and rotate the cable 180 degrees around the long axis. Or unplug both ends and rotate the whole cable around the long axis? Or try one end, then the other end, then both ends at once? Truly mystifying.

I'm particularly interested in the last paragraph. Is there anything out there that can read these emarkers and tell me the capabilities of a given cable? I have so many USB C cables that I honestly don't know which support what, so it'd be nice if I had a way to definitively determine their capabilities.

USB has become an odd sort of universal standard where everyone's best bet is to stick with the cable that came with the device.

I follow with that result, we establish a causal $ reason for this mess - USBIF signatory manufacturers wanting to not compete with the aftermarket.

This is a complete and utter failure of standard setting bodies.

From a consumer perspective there is no discernible standard to speak of.

On the plus side the reddit spreadsheet (!!!) about these cables ffritz posted suggests most charge at either 60 or 100W which is fast enough so I'll just avoid USB C for data transfers entirely

USB-C needs a Nutritional Information-like label:

- Speeds supported

- Currents supported

- etc.

Does anybody know how USB4 will impact this? Will any cable that will be labeled as USB4 need to be full spec, or will there be a similar situation of cables with varying degrees of compatibility all labeled USB4?

If we assume that USB4 is basically Thunderbolt 3 that means full spec cables will be too expensive so they will definitely allow lower-spec cables. And if it keeps using the USB-C connector then all the bad old cables are still there.

IIRC USB4 is planned to be exactly Thunderbolt 3.

What is needed is an easily identifiable branding, which I have some hope left USB4 will finally provide.

Another really shocking thing I recently discovered was the maximum length on a USB cable. USB 2.0 supports a max of 5 meters (~16 feet). Meanwhile a USB 3.0 cable only offers a max length of 3 meters.

And since USB 4 will be Thunder Bolt, expect cable length to become even shorter.

The standard absolutely needs to support the ability for computer and peripherals to detect what kind of cable is connected ... then we need consumer level implementation at the os to tell people when their cable is insufficient for the task they are trying to accomplish — with this, bring on all the cost saving measures for cables.

Without it — fucking idiot standards committees can go rot in hell for the blight they wrought upon the world ...

> The standard absolutely needs to support the ability for computer and peripherals to detect what kind of cable is connected

The USB-C standard has always had that ability. There's a pair of pins on the connector dedicated to that.

On a basic USB-C to USB-C cable, the first pin of this pair is connected to the corresponding pin on the other end, and the second pin is left unconnected; on an USB-C to USB-B or USB-A cable, this first pin is connected through a resistor with a specific value to the VBUS or GND pin (which one depends on the connector type); and on a more advanced USB-C to USB-C cable, or on an adapter to something like DisplayPort, the second pin is also connected, to a chip which informs the computer about the cable's extra functionality (and also negotiates the alternate mode in case of an adapter).

The device on either end of the cable uses that pair of pins to: negotiate which end will initially be the "host" and which one will be the "peripheral"; detect when the cable is inverted; ask the cable about its capabilities; ask the other device to increase the power or change which device is supplying power; negotiate alternate modes like DisplayPort; and a few other functions. The only thing these pins aren't used to, as far as I know, is to distinguish between an USB 2.0 and a basic USB 3.0 cable; instead, the same method as the older USB-A or USB-B plug is used.

> then we need consumer level implementation at the os to tell people when their cable is insufficient for the task they are trying to accomplish

The need for that alone is a big sign that the standard is flawed from a user perspective. Ideally, it should be impossible to use the wrong cable for a task. Traditionally, this is done by making the shape of the connector different.

For anyone else who was confused by:

> A USB 3.0 capable USB-B plug was physically larger than a 2.0 plug and would not fit into a USB 2.0-only receptacle.

recall that it's this USB-hubby one: https://www.amazon.com/TNP-SuperSpeed-Connector-Bi-Direction...

Oh, don't forget the sort of weirdo micro-USB 3 connector: https://www.droid-life.com/2013/09/26/yes-the-galaxy-note-3-...

It is admirable to try to maintain backwards compatibility, but still it is pretty strange connector for mass consumer market.

The backwards compatibility of that connector on my S5 saved my ass numerous times. I really appreciate that design.

Then there's also OnePlus Dash charge, which looks like a USB power brick with a USB-C cable. However the phone negotiates higher voltage with special circuitry in the adapter, so that kind of takes away from the 'Universal' in 'USB'. At least I now know that it might maybe work with a superspeed cable.

I look forward to usb4, which will merge with thunderbolt and simultaneously add and remove complexity from this situation.

I have never encountered a C-to-C cable with only USB 2 support. Are they are really a thing?

Most are in fact. When I searched for "USB C cable" on Amazon just now, I had to scroll to the 16th result to find a non-USB2 C-to-C cable.

When I search for "USB C cable" on Amazon UK, 90% of the results are A-to-C cables. The first C-to-C cable is an Amazon Basics 2.0 cable, so you are right.

Absolutely. They're often used as charging cables, e.g. on cell phones and Apple laptops, where engineering a high-speed data cable of the length required would be costly, difficult, and unnecessary.

Apple calls them charging cables. The advantage is they are less expensive for longer lengths.

They're less expensive, they're more flexible, and if you want to go to 3 meters long you don't have to worry about signal integrity.

Yes and since they are very thin they are quite useful when you just want to carry a tiny charger (my choice is the Mu One / RAVPower GaN 45W same Navitas charger except the Mu One has nifty folding international heads) and a lightweight cable. AmazonBasics is a reliable brand for this.

If they aren't, they will be, once the value engineering kicks into full gear :)

This is one of the reasons why USB-C doesn't particularly interest me. USB-C doesn't provide any benefits that I value, but comes with an increase in complexity and potential for (sometimes physically damaging) error.

> How many kinds of USB-C to USB-C cables are there?

Too many!

Any number greater than 1 is too many.

Huh. This was unknown unknown for me–I didn't know didn't know. I knew how to check device ports for markings, but it did not occur to me that I must check the cables too.

Nice! The list of six items is a more concise read than https://blog.fosketts.net/2016/10/29/total-nightmare-usb-c-t... , which also raised some concerns about the profound differentness of many identical looking cables.

In my oinion all USB-C cables should be identical and negotiation for power etc should be done inside the connected equipment, not inside a cable connecting the two pieces of equipment.

However it is an excellent opportunity to cheaply embed spyware chips inside USB cables with markings that cannot be distinguished from the genuine Electronic Marker chips.

It is so sad how perfect USB C could of been, and then they had to make so many variants that aren't easy to tell apart from one another not immediately obvious. I hope they do some USB3 stuff where they start coloring cables on the end that conform entirely to a given standard.

Here are the types, with names provide by me, what they would be called if USB-IF was sane and friendly:

USB-C v2a3: USB 2.0 rated at 3A

USB-C v2a5: USB 2.0 rated at 5A

USB-C v3a3 and v3g1a3: USB 3.2 Gen 1 (5gbps) rated at 3A

USB-C v3a5 and v3g1a5: USB 3.2 Gen 1 (5gbps) rated at 5A

USB-C v3g2a3 USB 3.2 Gen 2 (10gbps) rated at 3A

USB-C v3g2a5: USB 3.2 Gen 2 (10gpbs) rated at 5A

It's so refreshing to read an article that starts out like a rant and ends with an actually pretty decent solution. I look forward to the day when all my devices can tell me if the cable I've plugged in is good enough or not

What I want is for every display to have a USB-C port that can power an external thumbstick PC while transmitting video and audio in the other direction. eGPUs already work more or less the same way, so the technology is there.

Much like TB3 ports that have a small lighting icon, all USB-C ports and cables should have an icon that explains what type of cable you are expected to plug.

It would not completely solve the problem, but I think it would reduce it considerably.

It has become beyond ridiculous. We need to standardized around one cable. I get it, we can't have government doing this, but I'm at a complete loss for who else could? IEEE?

USB-IF already has certification but they aren't willing to kill off non-certified cables (which are 99% of the market).

I wonder if this puts them at risk of losing the "USB" trademark because they're not actively defending it.

Given the mess they've made is the certification/trademark actually worth anything? If what cable works with what devices is going to be a crap shoot anyway then it's not like it's a meaningful seal of quality to end users.

Even if I have a good cable, what I find confusing is what to expect USB-C speeds to be between devices.

Say I mount my GH5 with my MBP. What can I expect the speeds to be?

This fails to include Thunderbolt in the discussion, which is a major omission.

So, if I always buy "USB 3.2 Gen 2 (10gpbs) rated at 5A", am I ok?

Apparently yes, unless you need Thunderbolt!

Make sure it's USB-IF Certified.

Not to mention thunderbolt 3 even!

At the risk of sounding like I'm making a parody of an XKCD comic -- https://xkcd.com/927/ -- there does need to be a new standard physical connection that's compliant with TB3 and at least 100W PD. Then USB-C can be for backward compatibility and the new connection (USB-D? TD3PD?) can be future-looking only: no legacy support.

You could call it RichPort since the cables will be over $30 and the ports will only by found on >$1,000 Macs. I remember the old Firewire days and it was certainly simple but I don't know if the ecosystem was better overall.

I don't honestly know why anyone would design a product that used 100W via USB. Even without the issue of dodgy cables, it just sounds like a bad idea and a solution to a problem that shouldn't exist. If I ever ran across such a product I would question their other engineering decisions and pass on the purchase.

This is used for charging laptops. Personally, I love having a single cable for power, USB, and DisplayPort.

You ever wonder what the purpose of keying connectors is? I'm curious, because it's one of those time-tested ways to ensure that things get plugged into the correct ports and avoid costly accidents. Really simple too.

Why on earth is a standard as ridiculously complicated as USB-C a better idea? What is the practical value of an "everything" connector when its spec is so complicated that it really doesn't support "everything" at all?

Unfortunately, using identical connectors for incompatible protocols — often hazardously incompatible (to equipment) protocols — is an equally time-tested industry tradition, and not only in more cost-conscious market segments.

High-voltage differential SCSI comes immediately to mind — one might imagine there'd have been enough margin and price inelasticity in the market for $200 cables used to connect rooms full of $20,000+ storage devices to support the development and manufacture of slightly different connectors to prevent damage by connection to then-ubiquitous single-ended SCSI busses, but I suppose the projections for increased sales or reduced maintenance costs (for the vendors) due to redesigning the connector weren't high enough to justify a decision to standardize on different connectors for single-ended and differential SCSI.

Similarly, given that USB-C was essentially a fait accompli on the drawing board, short of material risk of injury or death due to misuse of otherwise compliant cables, I'm not convinced any one player in the industry — or any group of players likely to collaborate on something such as this — has both the means and the motive to seriously suggest something better.

Why build a better mousetrap when the world can't afford to beat a path away from your door?

> identical connectors for incompatible protocols — often hazardously incompatible (to equipment) protocols

The main actual hazard with USB-C is that somebody solders the low voltage power pins backwards. What that has happened, it could happen to absolutely any type of cable, so it's not exactly a problem with the standard.

For low volume production items, the plastics are more costly to design than the electronics. Therefore laptop docking stations are expensive to manufacture, especially as they are in use by only a subset of users.

By converting to a docking station that requires a simple small plug for everything, that makes them much cheaper. Instead of making a huge chunk of plastic that the laptop has to click into.

That explains why laptops should have a standardized "docking station" port, not why that port needs to also support literally everything that runs on electricity with the ability to also channel high-current power in both directions.

What's your threshold for "high-current"? USB ports have been delivering about two amps for a decade, and most USB laptop chargers cap at three amps. Even the absolute max of five isn't much higher.

because when it does it's magic. you connect one cable to a dock and your laptop becomes a desktop.

Maybe I'm just old, but having worked in IT for pretty much my entire working life I've come to the conclusion that "Magic" is not a good thing.

it isn't the 'i've got no f*g idea why it's working' kind, it's the 'they've finally managed to get it working and it really does work' kind.

There's another phrase for "Magic": Single Point of Failure

I personally also really love that I can charge my phone and my laptop with the same cable (I have a phone with USB-C that uses PD along with a 2016 MacBook Pro). It makes cable management much simpler, and I find it invaluable that when traveling I only need to bring one single charger with me.

Yep, it's great - my laptop, headphones, tablet and phone all have the same charging port. Even more, I can use many of the same MacBook dongles on my phone (including HDMI and mouse) to connect it to TVs.

Most laptop chargers are 65W, some are 95W, I haven't seen one over 100W. USB-C Power Delivery can do 100W (20V, 5A). More than 100W for the 1 in a million user is not the target of USB consortium.

Gaming laptop chargers often approach 200 watts. So it's uncommon but a lot more than one in a million.

I don't know but I did a search:

"Power adapters for Mac notebooks are available in 29W, 30W, 45W, 60W, 61W, 85W, and 87W varieties." [1]

Maybe you'd want your laptop to power external monitors via the other USB-C ports?

[1] https://support.apple.com/en-us/HT201700

Sounds like the Boeing 737 MAX problem (without the killing part).

Really what were the USB guys thinking.

I had to 'fix' my inlaws computer today because some shovelware browser was autostarting, and had a banner across the top to change the autostart settings.

How am I supposed to communicate the finer points of USB-c USB 3 and all the various combinations to them, when I cant even keep track myself. USB was successful because it was plug and play, simple, easy to use. All the things USB3c isn't.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact