For those of you who remember back almost exactly a year, last winter Stac Electronics filed a suit against Microsoft, alleging that Microsoft infringed on Stac's compression patents (TidBITS #164). Read back for the entire sordid tale, but the upshot is that a jury agreed with Stac that Microsoft had infringed, and awarded Stac $120 million in damages for the past infringement. That may sound like a lot, but Microsoft sold a lot of copies of MS-DOS 6.0 and 6.2 in the last year and it's small change to a company with $2.3 billion in petty cash. Stac isn't entirely in the clear though, since the jury also ruled in Microsoft's favor in a counter-suit that Stac misappropriated Microsoft trade secrets, which was only one of seven of Microsoft's counterclaims. The jury awarded Microsoft $13.6 million in damages as a result of Stac using undocumented calls in an attempt to make Stacker 3.1 compatible with Microsoft's undocumented data compression interface in MS-DOS 6.0.
Needless to say, both Microsoft and Stac say that they were innocent of any wrongdoing in the areas in which the jury found each guilty. Just once, I'd like to see a corporate lawyer to crack in court and start screaming "Yes, we did it! We stole their code. And we loved every minute of it! Ha ha ha ha!" Microsoft of course plans an appeal, no doubt hoping, if nothing else, to drag the suit out long enough to prevent Stac from being able to keep fighting.
Stac is seeking to have unsold copies of MS-DOS, including those pre-installed on computers, recalled, and Microsoft has already released a MS-DOS 6.2.1, which removes the DoubleSpace utility.
There are two interesting issues raised by this case. Stac was found guilty of appropriating (that's stealing in legal-speak, I suspect) Microsoft's secret "pre-loading" feature used in data compression. Now, this might in part be related to the fact that Stac had access to a beta of MS-DOS 6.0 (at which point there are different contractual agreements at stake), but the larger question is if using an undocumented call in an operating system is a violation of a trade secret? It would seem not, since the features in an operating system exist in part to provide services to other programs and utilities. If a call is undocumented and you use it, you certainly can't expect help from the OS folks, nor can you complain when your program breaks. But are you stealing a trade secret?
Related to this is the question of Microsoft's monopolistic leanings - not only did they include an undocumented feature in MS-DOS 6.0 for private use (a standard practice), but when someone uses that undocumented call to compete, they sue. We're not talking about a direct competitor here, since MS-DOS's DoubleSpace isn't sold separately, and it's not as though Stac itself sells versions of DOS (although to be fair, I believe Stacker is bundled with some other versions of DOS). So I could see an argument made that claims Microsoft plays dirty with third-party developers like Stac. I don't know if that's necessarily illegal, but I'm sure the FTC will evince some interest.
The second issue raised is even quirkier. You may have heard some of the hubbub surrounding the issue of software patents, which can cover such seemingly basic ideas as "cut and paste between files" (IBM patent #4,674,040). Many of these patented ideas have been arrived at independently, but often still run afoul of often-ludicrous royalties demanded by the owners. Another issue is the way in which the U.S. Patent Office conducts its reviews - some claim that the Patent Office's review board is not sufficiently qualified for the task and that the review process is cloaked in secrecy, meaning that a company can be liable for royalties well after they have independently developed a technology someone else just patented.
I'm distinctly not as up on these issues as I might be, but I wanted to mention them before 15-Mar-94, when the U.S. Patent Office stops taking comments from individuals for use in determining how the patent process should be changed for software. You can email comments to <firstname.lastname@example.org> and if you want more information about why software patents are detrimental, email the League for Programming Freedom at <email@example.com>. I don't know of a source for arguments in favor of software patents.
The reason this comes up in relation to the Stac/Microsoft suit is that Stac president Gary Clow said that Stac had shown that patent protection is one way small companies can fight back against behemoths like Microsoft. That's an interesting argument, and it's certainly valid, although Stac has used its patents to beat up on still smaller companies in the past as well (Salient and Sigma Designs, over the DoubleUp compression board - see TidBITS #164), so patents are not inherently a Robin Hood weapon. In addition, there are companies that do nothing more than purchase patents for the sole purpose of licensing them and collecting royalties. I don't believe that's the intention behind patent law, which, after all, was designed to handle physical machinery with its attendant huge research and development costs.