Hacker News Reader

Comment by ctz | Hacker News Reader

NXP is the most paranoid and ridiculous company I've ever had the displeasure of dealing with. Even after signing two separate NDAs with different parts of NXP, they still wouldn't give us complete sets of datasheets for parts we were considering. I don't know how they can sell anything at all; it's impossible to use their parts in a design.

View on HN

I've found NXP a lot better than Broadcom. They're like getting blood from a stone unless you're doing massive volumes... At least a lot of NXP's are freely available (like for all their micros)

Genuine question: why do they do this? Is it to protect their customers from reverse engineers?

I'd love to know, too! FBI agents were flipped for $100,000, just cause they needed the cash. Escaped detection for decades. CIA agents were turned, too.

While I'd not be surprised to see corporate security better than the FBI's stuff[1], I find it hard to believe you can't just get agents on the inside for the amount of money involved. So secrecy seems so unlikely to be useful against the companies that can really benefit from that knowledge.

It reminds me of Nikon not documenting their white balance settings in their RAW format. Just stupidity all around.

1: A long time ago, I saw Microsoft Internal Affairs in action. They were fast and setup a sting, complete with their own undercover agents/turned insiders, in a matter of weeks from first tip-off. I'd imagine companies like pharma or chips, would have even more severe response. Whereas in the FBI, one multi-decade mole was actually caught once or twice, but talked his way out. At one point, he even led the investigation against himself.

The Police Chief didn't know he was double crossed by Internal Affairs all along! I would fucking love to see one of those type of Dirty Harry films done in a megacorporate setting.

Here's another, Microsoft Internal Affairs more recently in action:

http://www.nytimes.com/2014/03/21/technology/microsofts-soft...

I've worked with all these guys, and the most plausible reason I observed boiled down to anti-counterfeiting: in their view, data sheets are a map to creating register- and pin-compatible clone parts.

While the knockoff market is very real, it always seemed far-fetched that NDAing data sheets would be that much of a protection. You would think a sophisticated shop has enough relationships to get the sheets via a close contact, friendly supplier, etc..

It's unfortunate that such backwards thinking persists. At some point you'd think they'd realize that datasheets are also a map to creating register- and pin-compatible designs.

A couple of reasons like inadvertent patent violation, to prevent clones, more closer to customer. The problem with this method is it requires lots of sales and support staff and they all want to focus on the big customers with huge run rates. So if you are a someone with a big pocketbook they will wait on you hand and foot. Anyone else is just not worth the time.

Can you tell us what's the nature of this part?

IMHO NXPs datasheets were usually ok (for common parts, of course)

I had a similar experience getting usable documentation for NXP's RF tuner ICs. In the end, I had to do a ton of guesswork on top of the partial documentation they were willing to share to end up with a working design.

Which is funny because at one point I stumbled upon a web site that was selling layer-by-layer microphotographs of the chip I was using. They weren't much use to me, but for someone wanting to reverse engineer the chip I'm sure they would be a lot more useful than a complete description of the chip's external interface.

Mifare Desfire EV1 RFID tag[1].

[1]: http://uk.farnell.com/nxp/mf3mod4101da4-05-1/rfid-read-write...

Well, that's why. The security on those is a joke.

Yay security by obscurity.

Ah I can understand why so many NDAs (especially after the security breaches they have had)

Were you in a position to use a competitor's parts?

Reply on news.ycombinator.com