Software Patents
and Freedom of Speech

I suppose it may seem a bit of a stretch to include the concept of software patents in with a discussion of digital rights management and copyright that was originally spawned due to discussion of the Blank Media Levy, but it actually hangs together for a number of reasons:

Because software is the result of a creative process, it should be covered under copyright, not patent. Many legislative bodies either have already or are in the process of, making software patents possible.

My writing this section was spurred today by an article in Slashdot pointing to an older article by Phil Salin written in July, 1991 that puts this better than any I've seen to date.

The article got me to thinking about the differences and similarities between copyright and patent from the point of view of the creator, and the fact that the size of the "owner" of copyright and patent actually is what makes these legislative tools realistic or not.

In very broad terms, copyright works best for individuals and patent for large corporations.

Let's look at this in more detail:

Copyright comes into existence as soon as a work is "published" the first time.

Patent must be applied for and requires (at the least) the aid of an expert in patents in the creation of the documentation (claims etc.) as well as the adjudication of bureaucrats as to whether the patent being applied for is innovative and unique.

From these two points we derive the fact that Copyright is inexpensive and Patent is expensive; Copyright is a given and Patent you have to work to get.

As a software creator (mostly in my past, but I still succumb to the temptation every now and then ;) I know for a fact that the practicality of either myself or most of the companies I've ever worked for doing a patent search on the possibility that there exists a patent on some aspect of the various programs we've created over the years is non-existent. In fact, I'm absolutely positive that software I've been responsible for infringes on one or more patents!

The fact is, the patents I or my programmers might have infringed should never have been granted in the first place precisely because I or my programmers could and did re-create the patented code with little or no significant effort - thereby invalidating the concept that the patent was the result of "significant effort".

On the other hand, there are probably lots of times when I or we have created something that the Patent office in its infinite wisdom might have deemed to be patentable, both software and "business practice" categories. For example, our company, Wimsey, was the first to send our invoices exclusively by e-mail to our Internet customers (we were selling e-mail addresses so knew that our customers had one.)

We can also take credit for some of the concept of using the DNS system for spam blocking (RBL.)

The problem is, being very busy creating and fixing the world's brand new Internet facilities, we didn't (couldn't) take the time and didn't have the money to pursue our "rights" and get patents on these concepts. I expect that even if we had had the time and money, we wouldn't have done so anyway since the solutions to the problems we saw were "obvious" to us.

On the other hand, I owe my abilities to program at all to the fact that I have had access to others' works and have the right under copyright law to create "derivative works" using the concepts and taking them in different directions from those of the original author. I just can't copy a piece of code character for character for long stretches (tens or hundreds of lines) because that would be copyright violation!

On the other hand, I can rewrite the code, using different variable names and comments, yet achieving the same end. In literature this is done all the time and gave rise to such examples as the Harlequin Romance books.

OK- but what about code based on unique algorithms?

XOR cursor, Compression, Encryption, Indexing, etc.

Here's a thought. If an algorithm must be expressed in assembler (human readable machine language for a particular computer architecture) in order to accomplish its claims, then maybe it should be patentable (maybe).

The means a couple of things:

And what about "Business Processes"?

"One click" purchase, international purchase via network with multiple currency conversion, etc.

At the point where it became practical to do international e-commerce amongst strangers (the general public), the Internet industry consisted of literally a world full of software geniuses with background in all types of commercial products and services. To them, the process and concepts of applying their background to this now ubiquitous and wonderful medium were "obvious". They'd been doing such things in private networks for years - admittedly mostly between related companies, but the same things none the less. For one company, not even the first, just the first with a patent lawyer at hand, to claim that all the rest now owe them a royalty for such obvious concepts is "patently" ridiculous.

And if Software Patents are upheld, then what?

Well, if government caves in to big business (and I think I've shown that only big business will gain in the case of software patents) then we should lobby for a change in the term of the patent from the typical 20 years to something like 3 to 5 years - roughly the generational life of computer systems today.

We now get into "Richard's view of the role of government" and how we, the general population, must constantly remind them (governments and our representatives) of the reason we put up with them.

So the governments have to figure out whether they are working for big business or the population. Big business can afford to purchase lobbyists. Small business drives the economy, and the population is the final pocket out of which all government funding ultimately comes and to which they have to answer each few years. I wonder which way things will go.

See my WebLog at blog.pacdat.net for more writings on this topic.