3 ms·

I think this is a bit hyperbolic. Once you try to implement any standard in practice to billions of people & devices you're going to find the same constraints a

by npunt 3y ago

I think this is a bit hyperbolic. Once you try to implement any standard in practice to billions of people & devices you're going to find the same constraints and similar trade-offs everywhere. Manufacturers are going to cheap out, always. There's no way around the fact that goods can get on the market claiming a thing that it doesn't do (perfectly) in practice. Tech is littered with semi-compliance. I hate to kick the can down to 'let consumers figure it out' but there's hardly a realistic alternative, you just can't police ~10 billion cables sold a year across ~10 million retailers. I haven't heard of your first example anywhere actually happening, it sounds theoretical more than widespread in practice in 2023, 8 years into this USB-C thing. As far as cables with different capabilities, what is the alternative here? 1. Physics of signaling demand certain capabilities require higher quality / more expensive cables. Zero chance you get people or manufacturers lined up around the idea of highest common denominator (e.g. 40gbps capable $30 cable when consumer wants a $3 charging cable), and that's sorta moot anyway because... 2. Standards evolve, and we're all tired of playing the game we've played for 50 years in tech of new physical connectors every few years. It's either a) you have the same physical connector and different cables, or b) different physical connectors that briefly use the same cables before they become differentiated when standards evolve. The reality is 'it works but not as good' (speed or charge-wise) is a superior tradeoff to 'I have to buy a totally different cable' in the high % of use cases where 'not as good' is possible, which is stuff like 'charge my phone' or 'plug my printer/audio device/whatever in'. Honestly the only unforced errors I see with USB-C are consumer education related. USB versioning should be simple (3, 4, 5) not intentionally obtuse ("3.2 gen 2x2"), and something like iconography should be added to indicate premium cable capability (specific higher voltage & bandwidth capabilities) to avoid to play the plug in and try it game.

p1mrx 3y ago
USB-IF should publish a table of every possible cable/adapter/receptacle type, with a short alphanumeric code that can be printed or molded onto the connector. The set of permutations is too large to fit into an icon, but a lookup table allows for unlimited detail.
- crote 3y ago
  There aren't that many permutations. Power-wise it is 60W, 100W (deprecated), and 240W. Data-wise it is 480Mbps, 5Gbps, 20Gbps, 40Gbps, 80Gbps. A simple power/data rating like already specified[0] is enough for >99% of cables. [0]: https://www.usb.org/sites/default/files/usb-if_usb_type-c_cable_logo_usage_guidelines_final_20230818.pdf https://www.usb.org/sites/default/files/usb-if_usb_type-c_ca...
  - npunt 3y ago
    Totally it's a very solvable problem w not many variants. It's unfortunate the USB group chose these particular labels. They're so ugly that you'd never want one on a cable, and they don't reduce to small sizes well. It's like they were never designed to be put on cables, or they never consulted anyone who ever used or sold a cable as to whether they were going to use it. I'd have preferred something that could be depressed into the plastic mold of the connector with two icons, a bolt and a dot: - default (USB2, whatever min watts) is blank cable - 60w has one bolt, 100w+ has two bolts - 5gbps has one dot, 20 has two, 40 three, 80 four That way you avoid the uncleanness of labels and their eventual scratching off, while maintaining at-a-glance (and even physical feel) understanding of cable capability. It also scales to future improvements. Honestly I think this would solve like 90% of people's complaints with USB-C.
  - carstenhag 3y ago
    Except for the dozens of manufacturer-specific power ratings which are not PD-conform.
    - crote 3y ago
      I am not aware of this problem existing in practice. Chargers and devices of course have a wide variety of power ratings, but those are generally all compatible with each other. Just get a charger with a higher rating than the device and it works. When it comes to cables, pretty much everyone just uses the standard 20V 3A / 20V 5A / 48V 5A cables. The only exception I am aware of is Oppo's SuperVOOC.
sokoloff 3y ago
"I have to buy a totally different cable" is something that most of us with over 20 devices with USB-C ports on them have probably experienced, only it's worse than that as it's more like "I have to buy several different totally different cables and try to remember to return the ones that also don't work". For a time at work, our desks had different docking stations for employees with Mac laptops or HP laptops. USB-C "standard" notwithstanding.
- npunt 3y ago
  So whats your solution? We all get that this is imperfect, the question is can the solution be improved upon within the constraints, and if not, what constraints should change? Because USB-C does a lot and touches a lot of things, I think there's a tendency to place the blame on it (and specifically its cables) for any circumstance where 'plug does not work as expected', which I agree is a frustrating problem that isn't fully worked out. I bet your HP docks are like Dell power adapters, which only provide specific volts/amps and maybe use some proprietary negotiation and aren't great device-agnostic chargers. I've experienced this firsthand, and unfortunately the blame lines squarely with HP/Dell/your IT department. It's also a sort of 'stage 1' problem in USB-C's rollout, in that manufacturers are used to doing things their way (e.g. proprietary power bricks), and the feedback loop of customer expectations hasn't yet reached a point to pressure them to change. When customers say 'this sucks it doesn't charge my iPad', HP/Dell/whoever begrudgingly realize they need to up their charger game, and little by little the problem is solved. I think politics might be the best lens to see USB-C through, because USB-C will only work as expected when consumers are providing that pressure. Hence my point about consumer education and clearer labeling being the unforced error of USB-C: moving up the learning curve quicker will minimize these problems and shake manufacturers into offering the thing that meet user expectations and the promise of USB-C. Ultimately someone has to make these things though and manufacturers all have their own misaligned incentives. It's the job of standards to meet them where they are and provide a path to aligning their incentives with better user experience. The reality is it's herding cats. I haven't seen anything in this thread or other USB-C threads that has offered a solution that improves upon the spec itself that is realistically possible within the constraints USB-C is operating in. That's unfortunate but I think represents the idea that USB-C is, to paraphrase, 'the worst type of connector, except for all the others'.

by npunt 3y ago

p1mrx 3y ago

USB-IF should publish a table of every possible cable/adapter/receptacle type, with a short alphanumeric code that can be printed or molded onto the connector. The set of permutations is too large to fit into an icon, but a lookup table allows for unlimited detail.

crote 3y ago
There aren't that many permutations. Power-wise it is 60W, 100W (deprecated), and 240W. Data-wise it is 480Mbps, 5Gbps, 20Gbps, 40Gbps, 80Gbps. A simple power/data rating like already specified[0] is enough for >99% of cables. [0]: https://www.usb.org/sites/default/files/usb-if_usb_type-c_cable_logo_usage_guidelines_final_20230818.pdf https://www.usb.org/sites/default/files/usb-if_usb_type-c_ca...
- npunt 3y ago
  Totally it's a very solvable problem w not many variants. It's unfortunate the USB group chose these particular labels. They're so ugly that you'd never want one on a cable, and they don't reduce to small sizes well. It's like they were never designed to be put on cables, or they never consulted anyone who ever used or sold a cable as to whether they were going to use it. I'd have preferred something that could be depressed into the plastic mold of the connector with two icons, a bolt and a dot: - default (USB2, whatever min watts) is blank cable - 60w has one bolt, 100w+ has two bolts - 5gbps has one dot, 20 has two, 40 three, 80 four That way you avoid the uncleanness of labels and their eventual scratching off, while maintaining at-a-glance (and even physical feel) understanding of cable capability. It also scales to future improvements. Honestly I think this would solve like 90% of people's complaints with USB-C.
- carstenhag 3y ago
  Except for the dozens of manufacturer-specific power ratings which are not PD-conform.
  - crote 3y ago
    I am not aware of this problem existing in practice. Chargers and devices of course have a wide variety of power ratings, but those are generally all compatible with each other. Just get a charger with a higher rating than the device and it works. When it comes to cables, pretty much everyone just uses the standard 20V 3A / 20V 5A / 48V 5A cables. The only exception I am aware of is Oppo's SuperVOOC.

sokoloff 3y ago

"I have to buy a totally different cable" is something that most of us with over 20 devices with USB-C ports on them have probably experienced, only it's worse than that as it's more like "I have to buy several different totally different cables and try to remember to return the ones that also don't work". For a time at work, our desks had different docking stations for employees with Mac laptops or HP laptops. USB-C "standard" notwithstanding.

npunt 3y ago
So whats your solution? We all get that this is imperfect, the question is can the solution be improved upon within the constraints, and if not, what constraints should change? Because USB-C does a lot and touches a lot of things, I think there's a tendency to place the blame on it (and specifically its cables) for any circumstance where 'plug does not work as expected', which I agree is a frustrating problem that isn't fully worked out. I bet your HP docks are like Dell power adapters, which only provide specific volts/amps and maybe use some proprietary negotiation and aren't great device-agnostic chargers. I've experienced this firsthand, and unfortunately the blame lines squarely with HP/Dell/your IT department. It's also a sort of 'stage 1' problem in USB-C's rollout, in that manufacturers are used to doing things their way (e.g. proprietary power bricks), and the feedback loop of customer expectations hasn't yet reached a point to pressure them to change. When customers say 'this sucks it doesn't charge my iPad', HP/Dell/whoever begrudgingly realize they need to up their charger game, and little by little the problem is solved. I think politics might be the best lens to see USB-C through, because USB-C will only work as expected when consumers are providing that pressure. Hence my point about consumer education and clearer labeling being the unforced error of USB-C: moving up the learning curve quicker will minimize these problems and shake manufacturers into offering the thing that meet user expectations and the promise of USB-C. Ultimately someone has to make these things though and manufacturers all have their own misaligned incentives. It's the job of standards to meet them where they are and provide a path to aligning their incentives with better user experience. The reality is it's herding cats. I haven't seen anything in this thread or other USB-C threads that has offered a solution that improves upon the spec itself that is realistically possible within the constraints USB-C is operating in. That's unfortunate but I think represents the idea that USB-C is, to paraphrase, 'the worst type of connector, except for all the others'.