This announcement is much, much more than it seems on the surface. Read the fourth paragraph carefully. National Instruments has gone public as it sets the stage for a lawsuit against The MathWorks, and the results of this battle will certainly have major implications for just about anyone using engineering software on a PC.
It doesn't take long with the business section of a newspaper to realize that legal maneuvers play an extremely important role in the success of high-tech companies, and that a judge's ruling can change the competitive landscape overnight. Very topical is the suit that MercExchange has brought up against eBay. The firm contends that eBay is infringing on several patents that cover methods of creating and searching online marketplaces and auctions. eBay makes no effort to hide the fact that a ruling against it could bring "significant harm."
Such suits concern business methods for which the federal courts legitimized patents several years ago. Those of us who have been hanging around the high-tech industry a while also remember some of the classic court cases that dealt with the then-new area of software law. Of course, we all well remember the Apple vs. Microsoft legal battledid the Windows GUI copy the look and feel of the Mac desktop metaphor? (As you can well extrapolate from the present day, Apple failed in its efforts to stop Microsoft.)
Not long thereafter came the battle between Lotus and Borland over whether the menu structure of the spreadsheet program Lotus 1-2-3 was copyrightable. As preliminary moves, Lotus first dispatched with two firms, Paperback Software and Mosaic Software, that were selling 1-2-3 knock-offs. Then it took on Borland, whose Quattro had its own menu structure, but that nonetheless could interpret 1-2-3 macros. That was sufficient for the legal system to ultimately sway towards Lotus. One of the best summaries of software and copyright law I've run into is from a course given at MIT, www.swiss.ai.mit.edu/6805/articles/int-prop/software-copyright.html.
Well, brace yourselves because there's a biggie coming up early next year. Here we have National Instruments taking The MathWorks to court. I haven't seen the actual papers, but from what I have heard and now see in this press release, the crux of the issue is whether the SimuLink block-diagram programming environment infringes on the patents that protect LabVIEW.
Just as Lotus set some legal precedents, National Instruments has taken a similar approach. Some products in National Instruments' software portfoliomost notably DASYLabcame to be there because of legal proceedings, where the target firms chose to "negotiate" (i.e., be acquired) rather than go to court. In this way, National Instruments also tried to set some legal precedents. Other firms simply stopped selling their software, one example being Intelligent Instrumentation with its Visual Designer. Draw your own conclusions, but National Instruments never chose to take on what many would consider the closest direct competitor to LabVIEW, Agilent VEE (originally known as HP VEE).
Now somebody has decided they would rather fight than simply back down. There have been unofficial rumblings about legal proceedings between National Instruments and The MathWorks for more than a year, but since announcements at NI Week last August and National Instruments' most recent quarterly report to the financial analysts, we know for sure. National Instruments has booked $3.5 million in expenses for "patent litigation" in the first three quarters of this calendar year alone, so you know they're taking it seriously. I have no doubt that The MathWorks is every bit as serious.
One reason is that those two firms are on a collision course in making their flagship packages a universal electronic engineering tool, with National Instruments coming from the data-acq end and moving to enhance the program's analysis and design functionality, while the MathWorks has its roots in signal analysis and has steadily expanded into data acquisition, control systems, and other ways to bring its traditional off-line analysis tools into working with real-world signals and systems.
Note that both are pushing their packages as front-end design tools for DSP algorithm development and code generation. In fact, The MathWorks' president Jack Little spoke on their cooperative efforts at this year's TI Developers Forum. They weren't on the stage, but at the conference National Instruments announced that it is integrating LabVIEW with TI's Code Composer Studio development tools.
Both are working with Tektronix and the open Windows Series oscilloscopes; evaluation copies of both MATLAB and LabVIEW are loaded on the scopes' hard disks. Both are setting up their graphical programming environments as front ends to FPGA tools from Xilinx.
Further, both tout their ability to develop and run code for real-time control systems, National Instruments with LabVIEW Real-Time and MathWorks with its xPC Target line, based on SimuLink. The two are starting to get in each others' way, to put it mildly, in just about every area.
See a pattern emerging? Any doubts about a collision course?
So, MathWorks presents the first serious challenge to National Instruments' patents, and it looks like a real Texas-style shootout is in the offing. Some people believe that the patents as granted were far too broad and should be overturned. Others suggest that there were prior uses that invalidate the patents. ("What about LabWorkbench running on a workstation from Masscomp, later called Concurrent Computer Corp.?" asks a friend of mine.) We'll just have to let the lawyers and judges make the final determination.
No matter what, the implications will be enormous. If National Instruments wins, it will likely be able to position LabVIEW as the sole universal EE software package, adding functionality continually over the coming years. If National Instruments loses, several things happen. First, The MathWorks will continue to give National Instruments some very serious competition as both companies continue down the same path. Second, the dam will break, and not only would companies that stopped selling their graphical software get back into the businesses, hardware companies will start developing and selling their own graphical programming environments, and startup companies will again start to spring up with new ideas.
No matter what the ruling early next year, I expect that the loser won't waste any time exhausting any possible appeals. It could be years before the dust finally settles. Meanwhile, life in the EE software business is getting very interesting.