A future without spreadsheets?

Spreadsheets are hot. You may not know it, but it was the humble spreadsheet that ignited the IT industry. They're not the sexist of applications, but they are probably the earliest example of that holy grail of software developers, the "killer app". So, how can I say that software patents would have prevented this development? Surely nothing as important as the spreadsheet could have been stopped by patenting?

The first example of a piece of software that we would recognise as a predecessor of the modern spreadsheet application is an application called VisiCalc. The creators of the software have a website, you can learn about VisiCalc there. It has a lot of the features we have come to recognise: the grid-based data layout, being able to set rules and change them, formulas, etc. Of course, it looks a little dated now, but in 1979 - the first time it was demonstrated - it was pretty neat stuff.

Now, in 1979, the amount of software patenting was very small, even in the United States. It wasn't until 1981 that the crucial cases began to be heard, and even then, it wasn't until the late 90's that people really began to wake up to the fact that they could patent all types of software. You can read my brief history of American patents to see all the key points in this timeframe. But, all you need to know is that 1979 was pre all the important court decisions that made software patents a reality.

Since VisiCalc was not patented, the marketplace was wide open and soon strong competition arrived on the scene. As people began to run into the limits of VisiCalc, the next major step forward was taken by an application called "Lotus 1-2-3". This became one of the strongest selling software titles of all time, and generated revenues of $53 million in the year it was released, 1983. In a number of ways 1-2-3 was very similar to VisiCalc, in others not so. There is no doubt that had VisiCalc been patented, 1-2-3 would have infringed, massively so.

This competition continued for another decade or more, and continued to be strong. As VisiCalc was toppled as top dog of the burgeoning spreadsheet/office software market, so 1-2-3 was toppled when Microsoft Excel arrived on the scene. Excel arrived at a key moment: as common operating systems developed graphical user interfaces (particularly, Windows), users began to want applications that would run natively in the graphical interface rather than in the text-driven DOS system. Excel was the first application to fulfil this requirement, and took marketshare away from 1-2-3 as Lotus struggled to re-write the application as a GUI application. However, Excel was still very similar to 1-2-3: it was very easy for users to transfer from one package to another. Had 1-2-3 been patented (which would have been almost impossible, since VisiCalc would have existed as prior art), Excel would have infringed that patent, massively.

So, what could have happened if the original VisiCalc was patented? Well, certainly that would have killed development in the spreadsheet market dead. Lotus 1-2-3 would never have existed, nor Excel after it. It would have been difficult to design a package that worked in a similar enough manner for users to be able to use effectively, yet not infringe on a well-designed patent. Without the strong competition and rapid development in the market, the software would not be anything like as sophisticated nor universal as it is today.

Would patents have had any benefit in this market? Unlikely - it's hard to see how a market could benefit from less competition, and there is little doubt that patents would have removed competition. Lotus strongly protected 1-2-3 with all the legal means at it's disposal, including landmark copyright law cases to protect the user interface of 1-2-3. If there had been a patent available to them to protect the functioning of the software, they would have used it to keep rivals out of the market. You might ask why it would be wrong for them to defend 1-2-3: of course, it would not be wrong. However, copyright law would not block entry to the market: if you produce an original work, you cannot infringe the copyright of a competitor. This is not the case with patents, though - any number of function in 1-2-3 or VisiCalc could have been patented, any one of which could have crucial to writing a spreadsheet.

Lotus 1-2-3 were the first company to encounter patent problems. They were litigated against in a case called the "Pardo Landau" case. You can read the patent application for yourself; it's quite hard going though, and it's not immediately apparent how this patent might apply to spreadsheet software. The claim that was used in this case was that a system of cells who automatically recalculated their values when one changed was patented, and as anyone who uses a spreadsheet will know, this recalculation process is one of the tenets of the spreadsheet: if figures and formulas did not update themselves when you changed a value that they used in their calculations, the spreadsheet would be no better than doing things by pen and paper.

Thankfully, in this case, it was decided that the scope of the patent was not sufficiently broad to cover spreadsheets - if you read the patent yourself, you will see it's not obvious how it covers spreadsheets. But in fact, it wasn't that bad a shout - it's pretty close. It also pre-dated modern spreadsheets by some years. But... if these people came close to patenting the spreadsheet, could it have been done?

We can leave the answer to the creators of VisiCalc, the original application which set the market on fire. This must be a frequently asked question of them, because they have dedicated a whole page on the question of whether or not VisiCalc could/should have been patented. And, as Dan Bricklin says in the introduction, "Why didn't we patent the spreadsheet? Were we stupid?". The answer pretty much comes down to the fact that they perhaps wrote the software too early - had they written the software later, it was likely that they would have patented it. He even says, "Personally, I think that the fact that software patents started being granted so late in the history of programming ... will cause all sorts of problems for the software industry.", while also saying that today he feels obliged to patent software. I'm not sure whether he's pro- or anti- patenting software - I suspect that he would be pro-software patenting, had patents been introduced in a sensible manner - but he does admit the chaos that has ensued, and the fact that we really would be much better off without them.