I had a patent issued last year as sole inventor (software, but a legit invention), and I actually just found a patent doc at a company I'm working with earlier today (they have the patent number in their web app's footer). I was planning to scan through the diagrams.
To me patent text is strange compiled pseudo-legalese, but I have found some learning in the diagrams.
Knowingly infringing a patent can bring triple damages.
You can bet that the USPTO granted someone a patent on the linked list, or a common variant of a hash table (don't get me started on how the USPTO determines prior art -- they don't use common literature, they seem to just search the patent database). So if you accidentally read a patent on said linked list / hash table and use it in a product and you're sued for infringement, damages go up quite a bit. That's one reason companies discourage engineers from reading patents.
Back in the 1980s someone had a patent on "Network standard byte order". Didn't matter the order you chose (big or little endian, or inside-out like the PDP-11, or whatever). As long as it was a stable and common byte order used on the network, you infringed. The "inventor" was a friend of a friend, and reportedly he said, when the patent was granted, "I can't believe they gave that to me."
Yeah, patents are utterly busted.
Similarly, if you have a patent pending, you are required to advise the examiner as soon as you come across something that appears relevant. When I had one under examination by the USPTO I sent in a patent that I saw here on HN (which the examiner decided was irrelevant).
The point is that you have intent. Reading something and not recognising it as prior art is very different from reading something, recognising it as prior art and not disclosing it.
The rules you see about "don't read any patents" come from the professional paranoia which is universal to lawyers. It is a much tighter defence to say "it's impossible that we had knowledge and chose not to disclose, because we did not have knowledge, because we have an ironclad policy of avoiding reading patents" than it is to enter into detailed argument about who read or understood which patents on which dates.
The distinction is important. A corporate policy of "don't read patents" is not the same as the law. It is a pre-emptive defence tactic intended to deny a hostile litigant from using an expensive line of argument.
Again, I am not a lawyer, this is by no means advice. I just hope to convey that the law is generally far more comprehensive, sensible and subtle than telephone games give it credit for.
- Patents are supposed to "teach" things. Yet in every place I've worked, employees are prohibited or at least actively discouraged from reading patents, because of liability concerns. So they fail at teaching because knowledge of what's in a patent is actually dangerous.
- The first month or so of reading those patents was basically about learning a whole new language. The terminology, twisted descriptions and deliberate obfuscation again makes the "teaching" aspect of patents utterly useless. No one in their right mind says, "I'm going to go learn some new stuff, so I'm off to read a bunch of patents." Nobody does that.
Patents don't teach, they're a land grab, and are about as hostile to the actual "Progress of Science and useful Arts" as a sucking chest wound.