well, as noted elsewhere in this thread:

- The knockoffs *are* shit quality in this case, flaky as hell, causing FTDI to field calls for issues with hardware that isn't even theirs
- knockoff quality is "indeterminate" and -- worse -- without recourse (there is no way to trace back who did shoddy work and get them to make good on the problems they caused)

Well, as noted, it happens with watchmakers, clothing/accessory companies as well, etc., etc.

No, you fail at QC.

What you're describing is what most reputable manufacturers do - they do destructive tests on random samples of each (lot,shift,day,week,whatever), few from the beginning, few from the middle, few from the end. As noted elsewhere, I've personally seen trays and trays of 486SX/DX chips ripped apart by IBM for testing prior to those lots' "acceptance" for use in IBM branded hardware.

So after they receive the hardware (ie, after the shipment swap risk) they test what they received. If the product leaves their possession and comes back (ie, attached to a board), they test samples of THOSE (also, again, destructively).

That QC is baked into the COGS for the product in question.

It doesn't absolve you of a need to QC what you're receiving. It just decreases the likelihood of getting bad product.

So buy through their designated supply channel and QC what you receive to ensure its provenance. Ask the manufacturer for appropriate QC steps to ensure that you are receiving "genuine product" that they manufactured. Make sure there is suitable indemnification in your contract with the supplier so you have recourse if they slip in bad product.

Again - these are all problems that reputable manufacturers have solved decades ago. It's the fly by nighters who have the most difficulty.

You do realize FTDI sells direct, right?

us.sales@ftdichip.com

Ah, so just because you haven't installed a working driver for that new hardware you just bought, the hardware is "broken"?

No, of course not.

Contact the manufacturer of the hardware and ask them for the driver. If they cop out and say "use FTDI's", say "FTDI's driver doesn't work with your hardware, where's YOUR driver?"

Actually, yes. China's government is very protective of ensuring that buyers are getting what they pay for, because the gov't doesn't want to upset the goose that's laying a Fort Knox worth of golden eggs in their country. The Chinese government knows exactly which side its bread is buttered on.

Chinese manufacturer management personnel who've done this sort of thing in the past with food products have been fucking executed, so don't for a minute think that there's no teeth to such an indemnification.

Not at all. It's completely accurate. The chip is still very capable of performing the task it was designed and marketed to do.

Someone just needs to write a driver that will talk to it, and it can happily do it.

Then you control your own manufacturing. Lots of companies do that very thing. Or they have very strict contracts with their manufacturers to lay the financial burden at their feet if there is parts-swapping happening.

Reputable manufacturers solve this problem every day. It's the cheap-ass folks trying to go on volume versus quality that have the problems.

But in this case, the manufacturer didn't choose that 0x0403. The USB working group did.

And - as discussed elsewhere in this thread - reputable manufacturers have means of holding their suppliers accountable, both in terms of QCing products they source, and in terms of indemnification for defective merchandise.

No, I think that about covers it.

If Company A isn't demanding test-samples from every shift-run, they're doing their QC wrong. That's just CODB.

If that's happening, I sue the bejeezus out of the factory worker. Or I decide "the risk isn't worth it", and control my own manufacturing.

There are solutions to all these problems. They may cost more up-front, but that's -- again -- the market normalizing itself as it weeds out the cancer of fraud.