The US government might be relenting on the Eolas patent. It’s about time the USPTO admitted that the patent should never have been issued.
What’s wrong with the patent process? Three things: A low bar to entry caused by incompetence in the USPTO (“Hey guys, my patent on eating food with a beverage came through!”) combined with opportunistic zero-morals holding companies who acquire patents and take advantage of their monopoly to sue everyone in sight. The third element is the inability of the USPTO to admit they’ve been wrong in issuing a patent for something obvious. Everyone knows there are bogus patents out there, but what can you do? You pay the $25K, and it’s cheaper than going to court. Try to convince a jury (of peers? don’t get me started) that XOR is something obvious to a skilled computer programmer.
There are other things that are wrong, like the “laches” that make it possible to submit a patent for something like, say, matter transmission now, and keep updating the submission until the technology happens, whereupon you hold a monopoly on something basic with (bonus!) a recent filing date. Nifty, huh? This happened recently with a guy (well, he was dead, but the, uh, decent company who’d bought his patent was standing up for him) who filed an application in 1955 for a patent on computer vision, which was extended and re-re-re-re-extended until it was finally granted recently. Whereupon it was used to sue the makers of bar-code readers.
Somewhere I have a copy of the infamous “Network byte-order” patent. Yup, someone managed to patent the technique of shipping bytes around the world in big-endian format. A friend-of-a-friend, he’s reported to have said “I can’t believe they gave that to me.” Thankfully, this miserable example has expired, but plenty more have not.