Combining copyrighted code with a machine (patentable) does not render this code patentable

Summary: Clarifying the situation in Canada, where law firms try to convince engineers to pursue software patents using loopholes that can help fool examiners but not courts

FOR BETTER OR for worse, Canada's patent policy has rarely been covered here (this reference page about it was last updated in 2013^*). It's a large country, albeit not a large population, and patents from Canada tend to target the US, with lawsuits too being filed in the US for maximal impact.

"The software must be more than an algorithm," says this new article titled "Patents in Canada". It was published some days ago and here is the relevant part:

To be patentable, software must be properly regarded as falling under one of the categories of invention outlined in Section 2 of the Patent Act, namely:

“any new and useful art, process, machine, manufacture or composition of matter, or any new and useful improvement in any art, process, machine, manufacture or composition of matter.”

The software must be more than an algorithm (Section 27(8) of the Patent Act prohibits: “mere scientific principles or abstract theorem”). If the software is a method and the method, on its own merits, does not pertain to patentable subject matter under Section 2 of the Patent Act, the method does not become patentable subject matter by virtue of being carried out on a computer.

Bereskin & Parr LLP's Paul Horbal and Denver Bandstra, both of whom working for the patent microcosm (law firm based in Canada), are trying to sell patent maximalism in a journal for patent maximalists. From the relevant passages (covering Mayo and Alice):

One common form of intellectual property protection sought for mobile apps and software generally is the “utility patent,” more commonly simply called a “patent.” A utility patent is sometimes referred to as the “Cadillac” of intellectual property; it can confer unparalleled status to the owner when compared to other forms of IP. Like its automotive counterpart, however, a utility patent can be expensive to obtain. In addition, the eligibility of software-based inventions for utility patents has been a topic of considerable debate in recent years. Courts in the United States and elsewhere have attempted to establish rules (e.g. Mayo and Alice) for when software-based inventions are eligible for patent protection. These rules can be difficult to parse, and often require the assistance of a professional to decide whether it will be possible to obtain a utility patent. Finally, a utility patent requires the owner to publicly disclose the inner workings of their invention, and to eventually make it freely available for use when the patent expires.

The above says that "the eligibility of software-based inventions for utility patents has been a topic of considerable debate in recent years." Many people wrongly assume that just because they put software "on the Internet" or "in a car" or "on a phone" that suddenly makes it innovative and/or physical. As any person with background in computing can tell, that's just utter nonsense. All these loopholes for patenting software have long been lobbied for by the patent microcosm and companies like IBM. When such patents get tested in court (as happens habitually in the US) they typically perish, so why grant these in the first place? ⬆ _____ ^* "Since 2005," says swpat.org, "the Canadian patent office's non-legally-binding Manual of Patent Office Practice talks of "computer-implemented inventions" and says "an act or series of acts performed by some physical agent upon some physical object and producing in such object some change either of character or condition" and "it must produce an essentially economic result in relation to trade, industry or commerce"."