Software patents do not provide the benefits other patents do. Name a piece of software that would likely not have been written if it wasn't able to be patent protected. Now compare that with other industries like pharmaceuticals, textiles, chemical processes, etc. Software is different because it's more straight forward engineering than explorative science. If an implementation is obvious to anyone with the prerequisite knowledge, it's not patentable and there's really not much, if anything, in the software realm that meets that criteria.
Software R&D has mostly not been patented for many years because algorithm patents are effectively unenforceable outside of narrow contexts, so it is largely futile. Computer science R&D is almost universally treated as trade secrets now, which have proven to be effective and defensible in many more cases.
The consequence of this is that the state-of-the-art in many areas of software are not in the public literature and there is no trivial way to learn it. Ubiquitous deployment in the cloud greatly limits the ability to reverse-engineer the underlying architectures, data structures, and algorithms. This is notionally the situation patents sought to avoid, but the practical unenforceability of algorithm patents has made it the default outcome regardless of whether there are patents on software.
