The Patents Bill

There’s been a fair bit of discussion in some quarters over a two word change to the Patents Bill.

The intent of the bill has been that software not be patentable. This is the case in Europe, and there is a lot of evidence that software patents have been stifling innovation, not fostering it. Actual software code can be copyrighted, but under the proposed law change, software will not longer be patentable. This is welcome by most of the local ICT industry. It stops patent trolls and the like.

However some NZ companies such as Fisher & Paykel Appliances said that software embedded in an appliance should still be patentable. The select committee and the Government agreed to this.

The problem has been how to define it. The bill was going to refer to “embedded software” as being exempt but now the Government has said the definition will be:

A computer program is not an invention for the purposes of this Act … only to the extent that a patent or an application relates to a computer program as such.

I’ve not been able to get a clear feel as to how significant this change is, adding on the words “as such”. Several IP lawyers have said it is good, while the Open Source Society has said it is bad, and they have a petition against the change.

I tend to be in favour of clear law, and think that is what the Government wants. However lawyer and software developed Guy Burgess makes a good case that the new definition is less clear than referrring to embedded software. He blogs:

There is no definition of “embedded computer program” in the Bill, nor does the industry-supported amendment include one. So because it’s undefined in the Bill, that creates “uncertainty”, right?

Wrong. Guess what else is not defined in the Bill, nor by any other law in New Zealand, nor by any Court in New Zealand: “computer program”. So how can there be objection to the lack of a definition for “embedded computer program”, when “computer program” itself is not defined? There can’t – and the same applies to many other terms in the Patents Bill.

The fact is, it is perfectly normal for undefined or hard-to-define terms to be used in laws. 

So Burgess argues it is not a big issue to keep the term “embedded computer program”. He notes it is commonly used:

But it’s not just Fisher & Paykel Appliances who have no difficulty with the term embedded software. A quick search on Seek.co.nz shows there are currently 48 jobs using the phrase “embedded software”, including Embedded Software Engineers, Embedded Software Developers, and so on. So while lawyers, for example, may claim the term is uncertain, the evidence is that the software industry itself has no difficulties with it.

There are many other resources confirming the commonplace use of the term in the IT industry. For example, there are over 42,000 articles on Google Scholar discussing “embedded software”, including from authoritative bodies such as the IEEE andAssociation for Computing Machinery. And of course there have been numerous patentsinvolving “embedded software” already granted.

So how about the alternative proposed by the Government:

Any possible questions about the industry-accepted term “embedded software” pale into insignificance compared to the European-style “as such” exclusion recently introduced into the Bill (i.e. the clause 10A(2) that the petition seeks to replace). If you ever wanted an example of notoriously uncertain, problematic language, it is the European “as such” terminology.

But don’t take my word for it – here’s what the Institute of Patent & Trade Mark Attorneys of Australia said in their 2010 submission to the NZ Minister (and which was expressly endorsed by the Australian Federation of Intellectual Property Attorneys):

[The EPC approach] will still introduce uncertainty and confusion as to the patentability of computer related inventions… The approach in Europe has also led to considerable uncertainty and criticismamongst European industry, the European Patent Office and the Courts… the approach has been often criticised as being inconsistent andcontradictory to the EPC itself… In our view, adopting the EPC approach would create a significant degree of uncertainty as to how the section is to apply and which, in any event, is inherent in the EPO approach.

Burgess concludes:

The exclusion of patents for computer programs remains, but with an express carve-out to ensure patentabilty of inventions containing embedded software – not the embedded software itself. Consequently there is no need to determine whether software is embedded or not to determine whether it is patentable. No “uncertainty” arises: under the industry-supported amendment, no software would be patentable. It’s pretty simple. And it’s entirely consistent with the Select Committee’s unanimous recommendation.

On the other hand, the European-style exclusion (which as the Australian patent attorneys say is uncertain, confusing, inconsistent, contradictory, unpredictable, and costly) is entirely contrary to the Select Committee’s unanimous recommendation.

Burgess makes a good case. I’m not 100% convinced that the proposed change is as dramatic as some claim, and think having the European style exemption is still a big improvement on our current law. But neither am I convinced that the proposed change brings benefits, over the select committee wording, and may be detrimental to our local software industry.

I welcome any op eds from interested parties on this issue.