5 ms·

If the device doesn't have the USB logo, it means they didn't pay the USB implementors forum. There are plenty of devices which implement USB, but don't pay the

If the device doesn't have the USB logo, it means they didn't pay the USB implementors forum. There are plenty of devices which implement USB, but don't pay the license fees. It doesn't mean they just output garbage. I've also seen devices with the vaunted logo on them which failed spectacularly when it came to following USB specs. I'm not sure what point you're making regarding the infeasibility of detecting counterfeit chips. It's not like these chips were bricked by accident. FTDI is specifically going out of its way to exploit bugs in these clones to brick them. So it looks like they have the detection thing pretty well covered. You'll note that the opensource linux drivers works just as well with the clones. That's because they do not expend extra effort to make them not work.

mrob 11y ago
The point is it's possible to build something that does not use USB, but which uses an almost identical protocol differing only in the meanings of the PIDs/VIDs. If it's based off USB 1.0 then any patents have expired, and if you don't call it "USB" or "FTDI" there is no trademark violation. In that case you would be doing absolutely nothing wrong, but the FTDI driver would still falsely detect it as "counterfeit".
- mrob 11y ago
  Not sure why I'm being downvoted here. In the USA, deliberately using counterfeit trademarks is a crime, not just a tort (see https://en.wikipedia.org/wiki/Trademark_Counterfeiting_Act_of_1984 https://en.wikipedia.org/wiki/Trademark_Counterfeiting_Act_o... ). The standard for being found guilty of a crime is "beyond reasonable doubt", not just "preponderance of the evidence" as it would be for a tort. There is no way to electronically establish counterfeiting has occurred beyond reasonable doubt, because the completely legal scenario I described is electronically indistinguishable. It's not acceptable to go accusing people of crimes when you know you can't meet that standard of proof. EDIT: As a practical example, consider the case of I2C. This is a proprietary NXP protocol, and you have to pay to get slave addresses. So instead people use "Two Wire Interface", which is not I2C but is electronically indistinguishable. The only way to detect counterfeit use of I2C is looking at the packaging/documentation.
  - kefka 11y ago
    That's why nobody is addressing "FTDI USB serial compatible". That too is a thing, and the Chinese use 16 bit arm chips, and still get them out for about $1/ea. and then, if I use "compatible", am I still a counterfeit? And why does my device deserve to be destroyed?
  - Dylan16807 11y ago
    >you have to pay to get slave addresses You what? But there's only a hundred of them.
    - mrob 11y ago
      There's also 10bit addressing mode. I couldn't find a reference to address fees on NXP's current website, but as recently as 2013 they had a FAQ on their website mentioning address allocation fees: https://web.archive.org/web/20130401000000*/http://www.nxp.com/products/interface_control/i2c/faq/ https://web.archive.org/web/20130401000000*/http://www.nxp.c...
- ascorbic 11y ago
  If you're connecting something that speaks your non-USB protocol without giving it a way to tell drivers that it's not USB then you can't really complain if th drivers think it's usb and act accordingly.
  - mrob 11y ago
    I have no problem with drivers thinking it's USB. That's not the same as actively sabotaging it and making accusations ("NON GENUINE DEVICE FOUND!") they can't prove.

mrob 11y ago

The point is it's possible to build something that does not use USB, but which uses an almost identical protocol differing only in the meanings of the PIDs/VIDs. If it's based off USB 1.0 then any patents have expired, and if you don't call it "USB" or "FTDI" there is no trademark violation. In that case you would be doing absolutely nothing wrong, but the FTDI driver would still falsely detect it as "counterfeit".

mrob 11y ago
Not sure why I'm being downvoted here. In the USA, deliberately using counterfeit trademarks is a crime, not just a tort (see https://en.wikipedia.org/wiki/Trademark_Counterfeiting_Act_of_1984 https://en.wikipedia.org/wiki/Trademark_Counterfeiting_Act_o... ). The standard for being found guilty of a crime is "beyond reasonable doubt", not just "preponderance of the evidence" as it would be for a tort. There is no way to electronically establish counterfeiting has occurred beyond reasonable doubt, because the completely legal scenario I described is electronically indistinguishable. It's not acceptable to go accusing people of crimes when you know you can't meet that standard of proof. EDIT: As a practical example, consider the case of I2C. This is a proprietary NXP protocol, and you have to pay to get slave addresses. So instead people use "Two Wire Interface", which is not I2C but is electronically indistinguishable. The only way to detect counterfeit use of I2C is looking at the packaging/documentation.
- kefka 11y ago
  That's why nobody is addressing "FTDI USB serial compatible". That too is a thing, and the Chinese use 16 bit arm chips, and still get them out for about $1/ea. and then, if I use "compatible", am I still a counterfeit? And why does my device deserve to be destroyed?
- Dylan16807 11y ago
  >you have to pay to get slave addresses You what? But there's only a hundred of them.
  - mrob 11y ago
    There's also 10bit addressing mode. I couldn't find a reference to address fees on NXP's current website, but as recently as 2013 they had a FAQ on their website mentioning address allocation fees: https://web.archive.org/web/20130401000000*/http://www.nxp.com/products/interface_control/i2c/faq/ https://web.archive.org/web/20130401000000*/http://www.nxp.c...
ascorbic 11y ago
If you're connecting something that speaks your non-USB protocol without giving it a way to tell drivers that it's not USB then you can't really complain if th drivers think it's usb and act accordingly.
- mrob 11y ago
  I have no problem with drivers thinking it's USB. That's not the same as actively sabotaging it and making accusations ("NON GENUINE DEVICE FOUND!") they can't prove.