Here’s the thing.
USB-B has EXACTLY the same four wires and USB-A does at the other end. Just in a different shape.
So, it is entirely on the peripheral (printer or whatever) to connect the correct pin to the correct input.
There is NO reason this could not be done with a USB-A style plug. It is just a different shape with exactly the same wires.
Next, you will tell me you can connect a DTE to a DTE or a DCE to a DCE without a crossover cable. They both have exactly the same pins, so why not?
Yes, they could have done this. And if they had, you would need to use the USB-A → USB-A cable you are so determined to use.
They could have also made the power socket on the printer a female power receptacle. If they had built it this way, the only way to supply power to your printer would be a power extension cord with a male plug at each end.
In both cases, there is nothing electrically wrong with this, which is what you keep harping on.
But both of these setups would require the user to utilize cables that can either damage electronics if you connect two power sources, or have 120 V on exposed prongs, which is why they call power cords with two male ends “suicide cables”.
The fact that these two cables would work in this specific case is beside the point.
The problem is that in other cases (which users are sure to do), they can damage equipment if you connect two power sources together (like two laptops), or are positively dangerous (in the case of the power cord with a male plug at each end).
All of your posts in this thread seem to be ignoring the fact, stated many times by many people above, that an A-to-A cable is dangerous. Do you not understand why they are dangerous, or do you not believe that they are dangerous, or do you not care that they are dangerous, or what?
@Whack-a-Mole, in 1996 what do you think could have happened if you accidently connected two hosts together with a USB cable with two USB-A connectors?
Forget the printer for a minute. Just connect two computers with a USB cable that has USB-A connectors on each end.
What could happen?
Sigh. No it isn’t. It encodes the kind of device that sits at each end. It prevents illegal and possibly damaging connections that violate the way USB is designed.
As has been mentioned multiple times, a double ended USB-A cable is the digital equivalent of a mains suicide cable. People still do stupid things like decide to connect a generator to their house with a suicide cable. And people die because of this stupidity. Doesn’t seem to stop people doing stupid things.
The original USB design was engineered to address a whole range of problems. And there are solid engineering reasons for the choices made.
Sadly over the years a lot of idiots have looked at the USB cable and thought of ways they could misuse it. Usually these involve using it to provide power. And such idiots create products that violate the standard without understanding or caring. So we see standard violating cables made available that are not intended for use with actual USB devices. Nor will they work with any computer or peripheral system. They might however damage your computer.
Say some stupid printer company decides to violate the standard and make a printer with a USB-A socket. And they supply it with a standards violating double ended cable. Their printer will only work with that cable. And that cable provides no protection against someone connecting two personal computers together with the cable. Something that could easily damage one or both. Why would any company be so idiotic to want to do this?
Over the years the demands on computer interconnection have grown. USB was very cheap, both in terms of hardware and licensing. But with the cheap and cheerful technology came implicit limitations. Way back then Apple came up with the vastly superior FireWire system. But it involved an order of magnitude more complexity to implement and Apple wanted a lot more to license the technology. FireWire is a peer to peer network out of the box. It was hot pluggable, dynamically reconfigured, boasted huge transfer speeds and such niceties as clock distribution. It made a whole generation of digital audio and video systems possible. But nobody used it for low end peripherals. You didn’t buy a FireWire mouse, keyboard or printer. I do have a FireWire film scanner and audio interface from then.
Fast forward to today and FireWire lives on as the underpinning of Thunderbolt. And you still don’t buy a Thunderbolt mouse, keyboard or printer. Because it would be silly. USB in its pedestrian form is still exactly what is needed.
What was made apparent was that as other devices became more capable, and themselves became fully fledged computers, even the basic interconnection between devices needed to step up. So the underlying protocols of USB-C came into existence.
One of the important enablers of newer technology is the ability to create cheap interface devices that have arbitrarily programmable pins. If you connect a mouse to a computer over a USB-C cable, chances are that the mouse is still a USB 1.0 device using a design decades old. Just with different plug. The first thing the computer does is talk to it as if it is a 1.0 device. And when it discovers it is, the pins in the computers’ USB-C socket continue to operate exactly as if it is. But the option exists for much more advanced communication to occur. The same computer hardware could discover that a Thunderbolt device had been connected. It will now drive those same pins with a totally different protocol. All manner of possibilities then exist. A network connection could be brought up, direct access to a disk could occur. And so on.
Technology advances. But we don’t revisit and violate earlier standards. We maintain backward compatibility and support. And that includes not suddenly deciding that a design that prevents damage to equipment can be thrown under the bus because we dislike the size of a connector.
In a sense, the OP is right that it would have been cool to have the same connector at both ends, but the technology wasn’t there yet, or at least it wasn’t cheap enough. Fast forward 25 years and we now have USB-C, which accomplishes this. It requires smarter devices, probably on both ends, and those smart devices would have been expensive or large or both back in 1996.
Agreed. Nitpick: 30 years. Which is a friggin’ eternity in the IT biz. A 1990s computer is about like a wooden sailing ship. Or maybe a dugout canoe. An interesting historical relic, but not the way to do anything useful in a cost-effective manner today.
It wasn’t practical with 1990s tech to make intelligent same-ended cords and intelligent cheap-ass peripherals, so they went with physically different connectors to enforce the necessary endianness.
And as a reminder, the endianness of USB-A vs -B was never about the data transmission. It was about the provision of (low) power to simple peripherals like mice that had no other power source for their own primitive simple electronics.
Is USB-B unnecessary in 2026? Sure. But it’s a legacy standard that’s still supported. Because enough peripheral manufacturers still find that USB-B ain’t broke and doesn’t require fixing.
I originally posted this in the pit, but i don’t think i want to waste space in that thread. So I’m copying my post (which replies to sometime @Whack-a-Mole said there) to this thread, and deleting it there:
It’s up to the system to get it right. There are different paths to make that happen. USB-C is a complex system where the attached components all talk to each other and negotiate who will do what. It’s a marvel of modern computing, and usually works. Back when the USB standard was first introduced, the shape of the plug was part of the protocol to negotiate which component did what.
Could that USB-B jack be replaced with a USB-C jack today. Absolutely. But that would mean replacing a mechanically sturdy jack with an intrinsically fragile one, designed for situations where size matters. (Specifically, where small size is paramount.) That would be a downgrade on a device that can support a larger jack, and that’s often used in a way that leads to, for example, people tripping on the cable. Could it be replaced with a USB-A jack? No, not without violating the USB protocol, which no legitimate manufacturer wants to do.
I was thinking from 1996 to around when USB-C started really gaining momentum, about 5 years ago to my addled brain.
Good point. And I was thinking until now.
Either measure makes sense, but on further review, I like yours better. It’s the more valid measure of the progress of the technology. Which was the point we were both making: USB-B was a product of its era, and USB-C is a product of its era.
From the other thread that has become this topic as well, this is it, the explanation that will truly work!
The question: why not a USB-A on both ends? Surely, the peripheral could deal with that shape of a connector.
Yes! A peripheral with a USB-A connector would work, because it’s a peripheral and will draw power, or not, as needed, from the correct lines. That’s not what the USB-B was designed to correct. USB-B was designed so that people wouldn’t plug two HOSTS together, both putting power on the same wires, with the result of frying one or both of the hosts.
Imagine one of those wall wart USB-A chargers – some bozo takes a double USB-A wire and just plugs them into each other, not realizing one side was already plugged in. Whoops! Magic smoke escapes and now there’s two broken chargers, or maybe even a fire.
Or, someone decides it would be cool if they could network their 1996-era laptops by connecting the USB-As together. Whoops! No more computer.
So, yes, peripherals could deal with that shape just fine, but the other shape was so people didn’t do things like connect two USB wall chargers together or two hosts together.
Yes, it’s just wires, yes, the peripheral would be fine. No, the hosts wouldn’t be fine, or the wall chargers. Now, 25 years later, smart connections are ultra small and cheap and along comes USB-C, which accomplishes what the OP is after.
OK, cool, this thread can be closed.
Moderating
At this point, the thread has devolved into essentially this:
Whack-a-Mole: USB-A and USB-B are exactly the same signals, just a different connector. Why?
Reply: To prevent you from connecting two hosts together and causing damage.
Whack-a-Mole: But USB-A and USB-B are exactly the same signals.
Reply: Yes, to prevent you from connecting two hosts together and causing damage.
Whack-a-Mole: But USB-A and USB-B are exactly the same signals.
Reply: Yes, to prevent you from connecting two hosts together and causing damage.
Whack-a-Mole: But USB-A and USB-B are exactly the same signals.
Reply: Yes, to prevent you from connecting two hosts together and causing damage.
And so forth, in an endlessly repeating loop.
Either you are incapable of understanding what should be a relatively simple concept (in which case, continuing to repeat this loop isn’t going to serve any useful purpose), or you’re just trolling.
Either way, this is closed.