> It's like pointing to a guy walking his dog at the park after noticing a parked car with it's lights on, and saying "there's the owner in the car" vs "there's

> It's like pointing to a guy walking his dog at the park after noticing a parked car with it's lights on, and saying "there's the owner in the car" vs "there's the owner of the car". So here's a bit of nuance for me. I can parse "the headers of the packets" in one of two ways: 1) For each packet, the header of said packet (clearly referring to the IP layer header) 2) For each packet, the headers of said packet (ambiguous) "The header of a packet" is relatively unambiguous, and akin to your analogy. I would grant this kind of mistake to be rare, although I'm not sure we'd agree on how rare. Certainly not impossibly rare - more common than those spurious 3s in my own writing, and yet we see one of those quite clearly. "The headers of a packet" is more ambiguous. A packet has only one header. The header has options, so that could be what's being referred to. Or it could be more loosely referring to metadata, or headers of another protocol contained within the packet data. I could see myself using this kind of loose terminology quite easily. I'm not a network engineer, I don't do switch configuration, I don't have to get down and dirty with IP layer issues with MTU and MRU misconfigurations, so it's quite possible I'm more prone to this kind of mistake than you are. I'm just a programmer who's occasionally asked to dabble with layer 7 networking stuff, and mostly interact with this when I e.g. fire up Wireshark to debug why my Amazon S3 authentication isn't working, only to discover Unity's WWW class ignored the (HTTP) headers I wanted to send on iOS (explaining why it worked perfectly fine on Windows!)