Date: Wed, 18 Apr 2001 17:48:19 +0000 (GMT) From: Terry Lambert <tlambert@primenet.com> To: tedm@toybox.placo.com (Ted Mittelstaedt) Cc: rsidd@physics.iisc.ernet.in (Rahul Siddharthan), djohnson@acuson.com (David Johnson), freebsd-advocacy@FreeBSD.ORG Subject: Re: Windriver, Slackware and FreeBSD Message-ID: <200104181748.KAA17086@usr02.primenet.com> In-Reply-To: <007201c0c7e1$65489b00$1401a8c0@tedm.placo.com> from "Ted Mittelstaedt" at Apr 18, 2001 01:27:24 AM
next in thread | previous in thread | raw e-mail | index | archive | help
> >Consider the following scenario: Apple has a patent on some very > >low-level algorithm, but doesn't tell people. (They do claim a patent > >on theming, so why not on some OS-related thing?) Their people (no > >doubt well-meaning) contribute it to FreeBSD. > > The second that an Apple employee formally contributed patented source > to FreeBSD, it would tremendously weaken the Apple patent to the point > where it would impede it's enforceability. Not really. What weakens a patent is distribution of an embodiment of the patent under license, since it grants rights to use the patent, unless they are explicitly disclaimed. In other words, if I put a patented algorithm (let's call a spade a spade here, and not try to call it a process) in code, and stuck a particular license on it, if I had the rights to do that with the patent, it would weaken the patent by the amount that the license permitted use of the patented algorithm. If I didn't have the rights to do that, and was acting on my own, I've granted absolutely no license. For example, SQUID uses algorithms which IBM claims to have patent rights to, and IBM goes out of its way to not ship SQUID with any of its products (going so far as to make Whistle rip it out of the InterJet II and replace it with Apache's cache). The Apache cache contains some of the same algortihms, but because there is no grant of license to the source code when you purchase the product, there is no grant of license to use the patent, and IBM is OK with that. > >It gets into -current, > >then into a stable release, becomes well entrenched into the OS. Then > >the legal people at Apple decide that FreeBSD has no right to > >distribute this patented stuff for free, and threaten to sue. > > It would take a couple years before becoming well entrenched, and if Apple > waited that long before doing anything about it, the patent would be > virtually unenforceable. Anyway, FreeBSD already went through this > with AT&T. A patent is not like a Trade Secret. It is enforcible until it expires. An historical common practice was to file for a patent, but not execute on the filing until someone came along and actually figured out a way to make money from what was patented. When that happened, the execution occurred, and the patent was then good for 7 or 14 years from date of issue. This was expanded to 17 years by treaty with Europe and other countries, which have absurdly long patent durations (a gamble that they can come up with a basic technology, and regain their status as superpowers, apparently). Such patents are called "submerged patents". They don't surface until someone starts making money off the idea, and then the company that surfaces them demands a cut of the profits (they start by demanding all of the profits, and negotiate from there). Recently, the term of patents was changed to 20 years from date of filing, rather than 17 years from date of issue, in an attempt to "deal with" submerged patents (once again, Europeans caused the length of term of a patent to be extended). These days, companies just keep information to themselves (the U.S. allows patents upt to 1 year following publication, but no more than that, and by treaty, the WIPO respects these delayed patents internationally). When someone starts to make money, that person usually files a patent, then the company that had the information does the same, and demonstrates "prior art", thus stealing the patent via trade secret prior art, so they really haven't gotten rid of submerged patents, what they've done is forced companies to be less forthcoming in their data publication (exactly the opposite of the intent of the basis of intellectual property law in the U.S. Constitution). In any case, what this boils down to is that they could sue anyone using FreeBSD who happened to end up having deep pockets, and collect for any use of the patented algorithm up to 20 years minus one day from their filing date, unless you could demostrate that they themselves licensed the patent to you. P.S.: The reason for the nominal fee on the source licenses from SCO and Sun is to ensure that the license is valid and limited by contract, since a valid contract requires the exchange of consideration. This is the same argument that UCB used to yank the Net/2 distribution as part of the USL/UCB settlement agreement, since if you did not pay for the distribution, there was no consideration, and without the consideration, they could void the license, even if there were no explict contractual clause permitting voiding. P.P.S.: This is also why Apple insisted on paying a token amount to UCSD for rights to use their P-system. When UCSD decided that it had value and removed it from distribution, they were unable to voide Apple's license, since consideration had been given to UCSD for Apple's license. Apple still has the only non-revokable license to the UCSD P-system. Unless you count JAVA as the successor to it, of course. Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-advocacy" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200104181748.KAA17086>