Imagine you choose a technology to base your business on, and some organization you're only dimly aware of sends you a letter saying you are infringing their patent or copyright.
The press has gone wild covering the SCO vs. Linux story:
C|net: http://news.com.com/2100-1016-991464.html?tag=nl
http://news.com.com/2100-1016_3-1017719.html
ComputerWorld: http://www.computerworld.com/news/special/pages/0,10911,2046,00.html
InternetWeek: http://www.internetwk.com/breakingNews/showArticle.jhtml?articleID=10300353
eWeek: http://www.eweek.com/article2/0,3959,1088440,00.asp
InformationWeek: http://www.informationweek.com/story/showArticle.jhtml?articleID=10300402
AlwaysOn: http://www.alwayson-network.com/comments.php?id=514_0_3_0_C
Robert X. Cringely: http://www.pbs.org/cringely/pulpit/pulpit20030529.html
Onlamp: http://www.onlamp.com/pub/wlg/3252
But this type of thing could happen in the VR world as well.
The VR world is smaller than the Linux world, but I've noticed a trend: every other person who talks about their equipment says that they just got an 8mm fisheye lens, and are using software tools x, y, or z to make their panoramas. None of which are products from the companies with patents covering making 8mm fisheye images into spherical panoramas. [6/17 note: perhaps after today's news gets resolved, there will only be one company offering a non-infringing panoramic stitching application].
Now I'm not a lawyer, and am not qualified to give legal advice. I'm trying to apply common sense to this highly charged situation. Consider what could happen if any of the companies with a fisheye patent decides to do a SCO to those photographers stitching with 8mm lenses and infringing software: they might send warning letters to all your customers; and bring you into court.
While all patents are supposed to describe inventions that accomplish tasks, they are informal groupings such as business process patents like the Amazon One-Click system; software patents like the RSA encryption patent, pharmaceuticals, and many other categories. Software patents are relatively new and much harder for the patent office to discover prior art than other categories of inventions. One reason for this is that software patents are often cloaked in the guise of a hardware device in order to get around the restriction that you can't patent an idea or algorithm. And there are literally thousands of places to look for software prior art, many of which may not be easily accessible or part of a patent examiner's process.
The history of such cases doesn't provide a clear guide. Consider UNISYS, which famously waited years before asserting their LZW patent over people writing out GIF files. They were ridiculed far and wide, but essentially won. Any software developer who provides GIF export (thereby using the LZW compression) in an application has to pay UNISYS or Compuserve a royalty. The PNG file format was designed to provide a patent-free replacement for GIF but has stalled out because Microsoft Explorer does not provide full support for PNG.
Then there is the RSA patent. The RSA patent for public key cryptography was challenged by Phil Zimmerman's publishing of PGP. Older versions of PGP (up to 2.3a) were thought to be violating the patent on the RSA encryption algorithm held by Public Key Partners (PKP). There was no court case, and since 2.3a, PGP was provided with the appropriate licenses in place.
And the Compton Multimedia patent (read this Wired article for background). This was a case where the patent office clearly missed prior art. They themselves rejected the issued patent after Compton's claims became famous and the patent office did more research on prior art in the multimedia field.
The conclusion of Christopher Chan's paper on Patent Rights on the Internet and in Cyberspace is a good guideline: "The Internet user should be aware of patent owners' rights as these rights are no less important than the commonly encountered copyright and trademark owners' rights."
According to Simson Garfinkel's article (see above Wired article), the average cost of a patent suit in 1995 was about half a million per side per claim. Simon Tatham states that "Patents work well in an environment where everybody has money. It costs to obtain a patent, and it costs to defend one, and it costs to challenge one." Unlike IBM in the SCO case, most photographers do not have the resources to withstand such a challenge. Yet hundreds of photographers are willing to play Russian Roulette with their financial future by making panos using 8mm lenses and infringing software.
If you lost in court, imagine having to reshoot everything you've done with a non-infringing lens. Imagine having to explain the whole mess to all your customers. Imagine the daunting legal costs. Choosing to violate someone's patents is not something that should be done lightly. At minumum it would seem prudent to consult an intellectual property lawyer familiar with patent and Internet law in your country, armed with the patents in question and your business plan. Do you really want to be the test case?
There are some products like the Panoscan which use an 8mm lens in a novel way: the machine acquires a single line of pixels through the lens at each step. According to comments I've heard, this process is quite different from that described in the fisheye patents, hence they have not been sued.
So my opinion is -- the safest bet for shooting panos with an 8mm lens is to work with software from the patent-owning companies. If you don't want to deal with the fisheye patent-owning companies, there is another solution which doesn't make you into a patent infringer: just use a different lens.
Addendum: the reader may find this documentation of a prior lawsuit involving the one of the fisheye patents interesting.
6/19 update: this article at KnoxNews.com has a few new quotes.
12:30:08 PM 
|
