Slashdot Mirror
IPFilter Infriging on Bay Network Patent?
jorhan writes "Darren Reed,
the author of IPFilter,
recently posted this message to the IPFilter mailing list. Apparently IPFilter may infringe upon USA patents owned by Bay Networks, specifically, #5790554. The patent might seem to own just about every conceivable way one might wish to filter and forward data packets, but trying to read through all of the "wherein said first condition" started to give me a headache (ObIANAL). But when you read what application the authors specifically had in mind, it really has little to do with network layer firewalling. Even more important is the question Darren's mail indirectly poses, "Anyone know of any prior art?""
I suppose the title could be talking about early adopters of Microsoft products, but that's not quite my intention. :)
The company that thinks of an idea that may be used widely later has the responsibility to patent it. The younger the technology is, the easier it is to get away with un-necessarely broad patent language, because people aren't aware of the number of uses that can fall within a patent's grounds.
My official recommendation for the situation is that tech patents granted in the last 5 years be reviewed by a panal of experts...patent lawyers from the government (FTC, department of commerce), paid consultants, and computer professionals from promonant comporations, i.e. Cisco Systems, IBM Corp, Microsoft, etc. to review their scope and reword them if necessary.
Note that it wouldn't be a party to get rid of tech patents, but to refine the existing ones as to nail down exactly what's protected and what isn't.
Bay Networks owns the patent, and as such, it is their responsibility to enforce it. Now, if they're nice, they could grant the authors if IPFilter a royalty-free license to use their intellectual property, but because IPFilter is an open source project, that is in effect granting the entire world permission to use it, and that is something Bay doesn't want. Hence, they need to stop the entire thing.
Conclusion: Yet another example of the shortcomings of the United States patent system. Sure, it's better than anything else in the world--but that doesn't mean it's perfect. Far from it, infact.
"Conclusion: Yet another example of the shortcomings of the United States patent system. Sure, it's better than anything else in the world--but that doesn't mean it's perfect. Far from it, infact."
Hmmmm, I'm curious - it's the "best" patent system in the world, but here we have "yet another example" of its failings.
How much do your actually know about the patent systems of every other country in the world?
In case you didn't already notice the patent office is in a pretty sad state, they will accept patents on virtually anything. This has resulted in companies filling for tons of frivolous patents on completely obvious technologies. That way if one of them tries to go after another for patent infringement that company can retaliate with it's own patents. The big looser in all of this is of course basically any non-corporate entity. Without a mile high stack of patents they become easy targets. I hate to sound naive but I'm a bit surprised at how little attention this has drawn in the political arena, you'd think by now someone would have started pushing for some reform but I haven't heard of any serious efforts to do so.
From the patent: the present invention relates to a method and apparatus for controlling the forwarding of data packets from a network device...
Seems obvious to me that this would affect a broad range of devices from switches to load-balancers to firewalls and would probably benefit a large group of corporations to begin either investigating prior art or ask Bay nicely to license the tech.
Reading the patent, both the abstract and the claims say many things to indicate that this patent covers network devices "such as a switch". Much of the patent is faily specific to forwarding between ports on such a device. I really don't think it can be said to generically cover generic layer-3+ packet filtering (in fact, I think it's pretty specifically layer-2ish).
Now, I'm not a lawyer, but I am a network engineer who deals with packet filtering all the time, and any "expert witness" worth his salt would bring these points up in a patent-suit. Someone should step up to be first on this one (Checkpoint or Cisco would be good choices, but there are many others who would be hurt by having to license this stuff).
On a more general point, I'm sure there are patents out there on just about everything that a modern Linux, BSD, etc system does. Some are already expired, but many are not. We really need to get a game plan here. My personal take is that patents are still a good thing, even on software, but it's the duration and disclosure that kill us. How can we reasonably get patent duration for software down to 2 years and require early disclosure of a pending patent? If those two things happened, patents would actually be a good thing for Open Source!
"My official recommendation for the situation is that tech patents granted in the last 5 years be reviewed by a panal of experts... patent lawers ... and computer professionals from promonant comporations, i.e. Cisco Systems, IBM Corp, Microsoft, etc."
What if an expert (from one of the big companies) comes across a patent from a competing company that would make things easier for their own company business-wise? They could argue for it to be removed. Of course, the other experts could probably just veto that opinion, but the influence is still there. This also works the other way too; a representative from one company will be in a better position to defend their own company's patents.
I say keep the professionals out of it because their own interests will taint the process. To replace them, bring in university professors that have nothing better to do than to sit on this panel of review.
Buying a Dell computer is equivalent to dropping the soap in a prison shower.
Patents aren't evil by nature. PKWare owns patents that cover the way the inflate/deflate alogrithms work. PKWare also put them in the public domain. Or the RTLinux patent. He wasn't served with papers or told to stop doing what he is doing. IPFilter isn't exactly an unknown piece of code either. I'd assume it's not a problem. Companies don't want to test patents like those becuase they lose all the marbles when they don't win in court.
Stop reading these articles or filter them out.
Or, better yet, it would be nice if there were some lawyer out there who would just donate a few hours his time and settle it for us. It's funny how the reputation of lawyers has fallen so low that it doesn't even cross our mind to ask them to do something charitable anymore (the extremely small minority of social justice lawyers excepted -- Ann Beeson, I love you.) Whether this is justified, I don't know. But there's no profession I frown on more, and goodness knows I won't be going to law school anytime soon.
:)
And I'm a newspaper reporter
I think there is a world market for maybe five personal web logs.
In the old days patents applied to the invention, not the result. You can't patent "a method for fastening clothing" but you can patent a zipper or velcro.
So it seems to me that filtering packets is a result, not a method, and as long as IPFilter doesn't use Bay Networks' code or some slick algorythm patented by them then I don't see why this should even be an issue.
Coding Blog
They actually admit that it's a specific case of a generic idea!
ALL patents are limited by the claims for God's sake! This is just boilerplate inserted by the lawyer. In fact it is quite often that a patent attorney who is writing a fair number of patents may insert a bit of boilerplate that he uses as a sort of signature, becuase patent's do not contain an author designation anywhere on the document. This bit of non-informational text may in fact just be the author's encoded signature.
Linux is not a UNIX. Get used to it.
/proc filesystem, and I print from my WinNT box at work using enscript and lpr because the Solaris machines at work don't have enscript and I don't have root.
UNIX, much like pornography, has gone to a "I know it when I see it" thing.
From the strictest view, one might consider the only UNIX OpenUNIX from Caldera. This is where the UNIX brand name has finally ended up (ATT -> Novell (strangely enough) -> SCO -> caldera). It may die there, caldera's in bad shape. OpenUNIX is changing, becoming very Linux friendly.
Almost all UNIXes (Unices?) have a Linux compatibility layer in the kernel. So Linux is becoming the one all encompassing API, if not the one true UNIX.
UNIX came out of AT & T, back when UNIX was still a research project and they were friendly with educational institutions. BSD was a fork. SVR4 UNIX, the most common "base" variant was basically SVR3 with BSD stuff. FreeBSD/NetBSD takes on the spirit of that work. Is FreeBSD UNIX?
Darwin, The base of MacOS is a Mach Microkernel with FreeBSD/NetBSD. It will be the most distributed "UNIX" ever. Is MacOS UNIX? it's very NeXT based, which was a bastard offshoot.
Linux works like UNIX, has the same design philosophy. Is the only UNIX some folks will ever touch. You have weird hybrids of SVR4/BSDlike systems depending on where Linus and the Distro guys picked and choosed stuff.
I have Cygwin on my Win2000 box. I use a bash shell, have rlogin, gcc tools. Is Win200 Linux? I even have a
Hmm, is POSIX compliance mean UNIX? POSIX was supposed to be the one true UNIX standard. If so than the most POSIX OS is WinNT. MS had a POSIX subsystem, never really worked but was needed to satisfy government regs on OS purchasing. MS WinNT was the only OS ever to get POSIX certified, so it's the one true UNIX, from a point of view.
Don't call folks stupid on things that are just interpretation. I can say OpenUNIX, the *BSD's, or WinNT the only UNIX, depending on what my criterea are. Sayig your interpretation is the only one is just trolling for a flamewar.
Patents should be reserved for true innovations, not something completely un-novel such as this.