An excellent article on the defensibility of software patents.
The problem is trivial and obvious patents with abundant prior art. Or taking two new technologies and combining them in a completely obvious way.
For example, I hold the US patent on using a SQL database to store I&A (Identification and Authentication) information. Basically using SQL to store passwords. As a result Microsoft had to use an ISAM database for Active Directory instead of the obvious choice of Microsoft SQL Server. Stupid and wasteful.
I presume that’s a hypothetical.
The XOR patent is always a favorite. I filed an amicus on another, where someone had patented networked bingo. I compared it to patenting faxing legal documents after the invention of the fax machine.
It’s not clear to me that software will ever be in a state where patents make sense. Frankly, it’s been my view for decades that there have only been three original (e.g., patentable) ideas in the entire field. Everything else either follows to one skilled in the art, or is a re-hash of prior art in some other field (e.g., networked bingo).
Generalized Security Policy Management System and Method, 5950195, 9/7/1999, claims the use of a relational database for general-purpose I&A (including arbitrary access policy such as time-of-day, client computer member group, and logon method). It was the claim for the relational database that gave Microsoft the ‘out’ to escape the patent by avoiding use of SQL.
This is called a combo patent - one that combines two software technologies in a way that ought to be completely obvious to a freshman CS student. And yet such patents are granted routinely by the USPTO.
Are there any firewalls on the market that aren’t violating that patent? Heck, if ActiveDirectory would have violated it, so does the Apache Webserver, which has a module for storing names / passwords in a relational DB for validating traffic requests.
The patent examiners don’t have the necessary knowledge to evaluate patents anymore. Too diverse and complex. The whole patent system is a huge waste of money and time. At the very least, they should halve the period of patent protection.