One of the comments on this blog suggested that an invention is “software” if it weighs nothing. This is an excellent illustration about how illusory and ill-defined the term “software'”really is.
In the general sense, software may be instructions executed by a processor. However, the “invention” is generally a method or sequence embodied by those instructions. The invention is not the code, but the underlying idea, regardless of the specific embodiment.
Another embodiment may be a set of transistors, gates, and other hardware logic that perform the exact same method or sequence. With hardware logic, the hardware may not be as flexible as using a general purpose processor, but the exact method or sequence would take place.
The “software” may be stored on a computer readable medium. Granted, adding the “software” to the medium does not change the weight of the medium, but arranging a bag full of transistors into a useful circuit that performs a patentable function adds no weight to the transistors, either.
Beforehand, the computer readable medium was useless. After writing the software on it, it becomes very useful and valuable and patentable. Similarly, the transistors are useless until organized.
I fail to see how “weightlessness” is the defining characteristics of software, as was suggested by one of the commentators.
This quick illustration is but one way to show how utterly ridiculous it is to try to carve out “software” from being covered by patents. The biggest difference between patents and copyrights is that the patent protects inventions that have practical utility, while copyrights protect the expression of a work.
In essence, is the anti-software patent movement suggesting that their software might be pretty, but is ultimately useless? I would not suggest that as a rallying cry.