USB 3.2 Specification Published

[starshipeleven]

Your average customer is happy when things work after they plug them in.

Which is not anywhere near guaranteed by USB Type-C and friends, for reasons detailed above (and copy-pasted below)

If you think the customers don't love it, but they would want more types of ports and with it more types of cables then you've missed an important trend in the IT.

No. I was talking about alternate-mode-capable Type-C ports being different from non-alternate-mode-capable Type-C ports.

Like making a USB Type-C-Alternate that is a bit longer or something, just to not be physically compatible.

As it is, I can get anything from USB 2.0 ports (yes, I've seen that) up to USB 3.2 in a Type-C connector, this is completely retarded, there is no way to tell what the hell is that port without specs or testing.

[starshipeleven]

aaand post above for sdack blocked by vBullettin.

[sdack]

... there is no way to tell what the hell is that port without specs or testing.

When you want to know what it does then plug it in and type "lsusb -v".

If this somehow doesn't satisfy you then get yourself a cube and play with it as long as you need to:

[dwagner]

I wonder why they still call it "Universal Serial Bus 3.2" if it isn't really serial anymore (but using two parallel lanes).

[Adarion]

It seems USB is still having the growing pains. At least different colours or some hints could help to find matching cable/device/interface combinations. And that's just the outside... The USB3+ situation is quite a mess sometimes. Though older USB still has its share of problems. And if you have no old-school dedicated interfaces you have to connect everything via USB. So far that could be okay if USB wouldn't mess up. Attach more devices and more and troubles start. Power problems, enumeration, bandwith problems and so on.
Moreover it is overkill in terms of bandwidth for several applications like user input or some cases of machine driving - and in the latter case you might even get to feel the limitations of highspeed connections that are susceptible to disturbances, long cables are usually not possible with USB. Serial was kinda robust on that.

[Delgarde]

Your average customer is happy when things work after they plug them in.

The problem is, they don't work, because the user can't just rely on "the plug fits, it'll work" - they need to know if that port is capable of DP-over-USB, and if it's not, they have to figure out why things aren't working (maybe keep trying other USB ports until you find one that works, or try to decipher the little icons next to them).

[sdack]

The problem is, they don't work, because the user can't just rely on "the plug fits, it'll work" - they need to know if that port is capable of DP-over-USB, and if it's not, they have to figure out why things aren't working (maybe keep trying other USB ports until you find one that works, or try to decipher the little icons next to them).

No. This is just a common fallacy, which you will always have. People have always been plugging things in and then wondering why it's not working. It's not possible to make this absolutely fail-safe, because you just cannot eliminate the physical world. You can always get some dust and dirt in between or have a cable break. However, you can reduce the amount of problems by avoiding different shapes and avoiding intentional incompatibilities.

Usually what comes next is the user trying to figure out if the driver works and if the protocols are compatible, and finally if the software is working as expected. This is the part you want to happen automatically. Ideally should your OS be telling you what you've plugged in, what is working and what isn't, or if there are any problems or only inefficiencies with the connection.

Do remind yourself that in the past you often couldn't make a connection at all and that the problems you're seeing now are just a matter of "the glass being half-full" while you see it as "half-empty". You can certainly implement a feature, which will not only show you with your notifications what you've plugged in, but also where it's plugged in, what speed it's currently running at, what features it supports, and even tell you if there is a better port to plug it into. This do I see as more useful than knowing that a green cable doesn't fit into a red port.

[patstew]

Which is not anywhere near guaranteed by USB Type-C and friends, for reasons detailed above (and copy-pasted below)

No. I was talking about alternate-mode-capable Type-C ports being different from non-alternate-mode-capable Type-C ports.

Like making a USB Type-C-Alternate that is a bit longer or something, just to not be physically compatible.

As it is, I can get anything from USB 2.0 ports (yes, I've seen that) up to USB 3.2 in a Type-C connector, this is completely retarded, there is no way to tell what the hell is that port without specs or testing.

So you think that having to have USB-C to USB-C, USB-C-Alt to USB-C and USB-C-Alt to USB-C-Alt cables in addition to all the cables that already exist would improve matters?
The fact is that a flash drive is never going to make a video signal when you plug it into a TV, a Raspberry Pi is never going to provide 100W to a downstream device, and a keyboard isn't going to stick 20Gbps transceivers on its circuit board. No amount of interface standardisation is going to fix that, it's just the nature of the devices in question. Making everything work with one cable and connector and allowing multiplexing is a big improvement though. I have one USB-C cable that I plug into my laptop at work that connects an extra monitor, mouse, ethernet, sound, USB stuff and charges it. And it's standard so it also works with my phone, and with the next laptop I get, unlike a proprietary docking station. The convenience and reduction in the variety of cables I need massively outweighs the fact that it's possible to physically connect two devices that don't work together.
Surely the way to solve your issue would just be to print the Thunderbolt/DP/HDMI logos near the port? Most laptops do this anyway, at least for Thunderbolt AFAIK.

[notanoob]

USB 3.0 was fine, 3.1 is the clusterfuck. Specifically, the voltage situation and the whole gen1/gen2 situation. Everything else about it is fine. But seriously, we don't need more bandwidth. I don't know of anything that fully utilizes the bandwidth in 3.1 gen 2, so, I don't think there's an industry demand for more than 10Gbps.

Actually if you are making a server farm out of GPU's having access to larger bandwidth over usb is very useful. Before you could only hotplug GPU's using thunderbolt and achieve reasonable performance. Now you don't need thunderbolt to use external GPU's.

This is also useful for those who make custom hotplugable laptop addon's out of old GPU's because the bandwidth allows for better gaming. But the real question is who could possibly want to hotplug GPU's since laptop gaming is a poor choice. Desktops make better gaming machines with a laptop for battery life and portability.

[Zan Lynx]

The problem is, they don't work, because the user can't just rely on "the plug fits, it'll work" - they need to know if that port is capable of DP-over-USB, and if it's not, they have to figure out why things aren't working (maybe keep trying other USB ports until you find one that works, or try to decipher the little icons next to them).

This is why Microsoft has error messages, as described here: Troubleshoot messages for a USB Type-C Windows system

Notice the ones about "DisplayPort/MHL may not work."

Hopefully someone working on Linux desktop environments will write some similar notifications, so that people are informed when an alternate mode isn't available, or if there isn't enough power. Or if their cable isn't working.