[fake-name]

I mean, the entire article is basically documentation of that research.

This isn't a peer-reviewed piece of work, it's a writeup of someone's fairly exhaustive research into a problem they encountered.

I don't see why you'd need confirmation from another person that something you bought doesn't do what it's part number claims it should.

[andi999]

Maybe I didnt read it carefully enough but to me he does not research it but just states it. I mean some (non counterfeit) production runs might not have the expected quality isnt this possible? If the rom signature is not garanteed by the ip holder than it still might be legit.

Probably he is right though, but to me it reads like the conclusion is the premise. (Might be due to the writeup though)

Edit: here is my bone: it says: how do I know? If the ROM does not follow the pattern 28-xx-xx-xx-xx-00-00-xx then the DS18B20 sensor is a clone [5]. And here I would have expected [5] to be the datasheet or something, but not 'own research'. The idea of citations is also to make your claims more verifiable.

Now, if we look at the Datasheet: https://datasheets.maximintegrated.com/en/ds/DS18B20.pdf it actually says:"The least significant 8 bits of the ROM code contain the DS18B20’s 1-Wire family code: 28h. The next 48 bits contain a unique serial number. The most significant 8 bits contain a cyclic redundancy check (CRC) byte that is calculated from the first 56 bits of the ROM code. A detailed explanation of the CRC bits is provided in the CRC Generation section." So the 28 is required. The '00-00' part is ust the higher bits of the unique serial number.

I wouldnt be surprised if different factories get different higher bits.