Who Invented the Linux-Based Wireless Router?

mtaht writes "I've just had the interesting experience of being deposed to talk about one of the first embedded, Linux-based, wireless routers. Our (free!) 1998 publication of how to make one predates patent #7035281, filed September 13, 2000, by someone else. Their patent was recently granted and is now being disputed in court, in part using our how-to as an example of prior art. The lawsuit continues; the case goes before a judge shortly, and a jury trial if necessary is scheduled for the spring. I find myself plagued with the question: So... who invented the embedded Linux based wireless router? What relevance does 'who' have, when there is such an enormous confluence of ideas from thousands of people? What constitutes invention, anyway?"

Comments

Finding useful prior art takes work

[ciaran_o_riordan]

Please read how to do it right:

http://en.swpat.org/wiki/How_to_read_patents_and_gather_prior_art

Re:Finding useful prior art takes work

[Shadow Wrought]

Did you even read TF Summary? He is the prior art! He's just trying to figure out if there was someone there before him, or how to give credit to those pieces which were already built into the Linux code.

Re:Finding useful prior art takes work

[mtaht]

Thank you! this is excellent, so far.

wuh?

[stoolpigeon]

I thought we just resolved this in the last story. Vint Cerf invented it and the lawsuits are his fault.

Re:wuh?

[bill_mcgonigle]

Thanks, a Slashdot comment hasn't made me chortle like that in a while.

Fails obviousness flow chart.

[kurokame]

Does it have Linux on it? --no--> Install Linux. Is Linux installed? --no--> Modify Linux until you can install it. ----> Install Linux.

It's Linux. Installing it is always obvious.

Re:Fails obviousness flow chart.

[tepples]

Is Linux installed? --no--> Modify Linux until you can install it.

That can prove difficult if a device won't execute a kernel until it has verified the RSA-encrypted secure hash of the kernel. Have you updated your flowchart to account for methods that have become commonplace to "secure" appliances such as home routers and pocket-size computers?

Re:Fails obviousness flow chart.

[h4rr4r]

Replace its boot loader, this may take hardware modification.

Re:Fails obviousness flow chart.

[tepples]

Good luck replacing a boot loader whose first stage is stored in ROM on the same die as the CPU.

Re:Fails obviousness flow chart.

[Tigersmind]

Does it have Linux on it? --no--> Install Linux. Is Linux installed? --no--> Modify Linux until you can install it. ----> Install Linux.

It's Linux. Installing it is always obvious.

Why is it I always agree with posts modded Funny?

Re:Fails obviousness flow chart.

[tagno25]

Replace CPU with one bought from Digikey, Mouser, or other large part seller.

Re:Fails obviousness flow chart.

[tepples]

Replace CPU with one bought from Digikey, Mouser, or other large part seller.

And then find that this CPU has not only a custom BGA pinout but also a GPU and other *thbridge-integrated components on the same die. So by that time, you could have bought yourself a BeagleBoard or reserved a Pandora PDA instead of modding an appliance.

Re:Fails obviousness flow chart.

[Runaway1956]

Replace CPU with one bought from Digikey, Mouser, or other large part seller.

So - we need a bigger CPU? Alright, alright, I had to intentionally mis-read your post to come up with that, but hey, BIGGER is always BETTER, right?

What constitutes invention?

[popo]

Every new platform (almost by definition of the term "platform") allows it to be married to myriad other technologies. Unfortunately the USPTO does not seem to understand that each one of these secondary permutations does not (should not) constitute "invention".

But ultimately -- the answer, like most answers to legal questions, isn't a matter of who's right and who's wrong, or even "what is the truth?". The answer depends on who has the combination of legal-budget and willpower to pursue the patent.

Re:What constitutes invention?

[gstoddart]

Every new platform (almost by definition of the term "platform") allows it to be married to myriad other technologies. Unfortunately the USPTO does not seem to understand that each one of these secondary permutations does not (should not) constitute "invention".

And, that's the problem with patenting some of these things. I fail to see how that patent should be allowed to stand.

Routers existed. Linux existed. Wireless existed. Hell, TFA sums it up very nicely:

My mental question remains. Did Greg, Everett and I really invent the embedded Linux based wireless router?

All we did, basically, was take code that already existed, compile a new driver, install a board, make a few cables, and prove such a box could stay running in a world where people trusted IOS. We're just the first people that bothered to plug in a wireless card into a junked PC, boot Linux off of a floppy, run wirelessly 13.1 miles and then publish how to make it work, in plain english, a howto a more general public, and even a patent lawyer, could understand.

Amusingly enough, our little howto hung off the far end of that wireless connection for years, dissipating electrons in the airwaves, for every one of the tens of thousands of hits we ultimately got. Everybody ate from our dogfood, in other words.

They didn't "invent" anything. They did do something new, and then they shared it like nice people. I just fail to see how putting together three existing technologies in what is a fairly logical configuration merits a patent.

I hope this patent gets dismissed. Of course, that would only be one of bazillions of patents which make no sense whatsoever.

Re:What constitutes invention?

[noidentity]

Woah, I just had this new idea for an invention: a Linux-based wireless router that has TWO power connections, instead of just one. I need to go patent this great, innovative idea right now.

Re:What constitutes invention?

[DamageLabs]

The patent office does not care about your patent. Never did, and never will.
It just cares about the patent application, its form and fees. It is up to the patent applicator to enforce the patent and ensure there wasn't prior art.

If there was, the spent money on the patent and the lawyers is a waste, but it does not come out of the patent office pockets. Actually, everybody makes money in a situation like this, except for the patent applicator.

Re:What constitutes invention?

[Machtyn]

I don't know... I think you'll be infringing on my patent of a Linux-based wireless router that has X-number power connections and Y-number antennas.

Re:What constitutes invention?

[NFN_NLN]

Every new platform (almost by definition of the term "platform") allows it to be married to myriad other technologies. Unfortunately the USPTO does not seem to understand that each one of these secondary permutations does not (should not) constitute "invention".

But ultimately -- the answer, like most answers to legal questions, isn't a matter of who's right and who's wrong, or even "what is the truth?". The answer depends on who has the combination of legal-budget and willpower to pursue the patent.

I think the confusion stems from Ben Franklin who invented the "Franklin Stove" by marrying two cutting edge pieces of technology: a stove and a fucking door.

http://en.wikipedia.org/wiki/Franklin_stove

Re:What constitutes invention?

[francium de neobie]

I'm sure you're infringing on my patent on a device that consumes electricity and sends or receives wireless signals.

Re:What constitutes invention?

[slinches]

They didn't "invent" anything. They did do something new, and then they shared it like nice people. I just fail to see how putting together three existing technologies in what is a fairly logical configuration merits a patent.

I'm not so sure it's that simple. Nearly every mechanical device invented has just been a combination of simple machines in different configurations. I think the distinction between invention and merely an alternative application is whether the item being patented provides a new capability or one that is significantly improved.

Using this case as a hypothetical example. If there were already wireless routers running a proprietary OS and the OS was just replaced with another that provided similar functionality, that would not (in my opinion) be an invention. If instead, there was no such thing as a wireless router at the time and they used Linux, an antenna and some existing computer hardware to create one, that would be a novel and patentable invention.

I don't know all of the details of the state of the art at the time or what the patent actually claims, but from what I do know, I don't think this patent falls clearly into either category.

Re:What constitutes invention?

[gstoddart]

You don't seem to know what 'invent' means.

Well, I don't know what the legal threshold for "inventing" is in this case. The guy who did 'invent' it isn't sure he actually invented anything. And he did it before the guy who patented it.

Built? Sure. Assembled? Fine. Extended someting? Absolutely. Invented? I honestly don't know.

Take any functionality that already existed on a computer in the late 90's, add "wirelessly" to it -- have you "invented" anything? Or extended something that was already well known? I would argue it's a (fairly) predictable application of existing tech -- wireless is just one in a long line of 802.* protocols; does going from 802.3 to 802.11 cause magic to happen?

They didn't invent the router. They didn't invent wireless networking. They just sorta smushed them together, and in a way that is consistent with how you might expect them to be used.

Did Jeff Bezos "invent" one-click purchases? Or did he basically take the well-known concept of "button" and apply it to the well-known concept of "purchase"? Many of us would argue that it's a stupid patent.

This is before the courts, and has to go through a jury trial. It might be a little premature to get all smug and say that you can definitively conclude that this was, in fact, an "invention" or not. If it was so easy, we wouldn't be reading the article and debating what exactly "invent" means in this case.

Re:What constitutes invention?

[gstoddart]

I'm not so sure it's that simple.

Nothing ever is, which is why we like to come to Slashdot to discuss it. ;-)

Nearly every mechanical device invented has just been a combination of simple machines in different configurations.

Yes, that's true. And I don't mean to say people never truly invent things because all machines can be reduced to the lever, wheel, ramp, or its other base components. I'm saying that sometimes when it's applied to software, one finds oneself looking at a patent that says "method for doing a well known task with a computer (or wirelessly, or over a network)". Have you "invented" something, or applied technology?

I don't know all of the details of the state of the art at the time or what the patent actually claims, but from what I do know, I don't think this patent falls clearly into either category.

Well, essentially the guy being discussed built this before the patent was filed for, and publicly told lots of other people how to do it.

So, if there was prior art, the patent is null and void. Or, did the guys with the patent actually create something which was non obvious?

Obviously, we here won't settle this conclusively. It's just sometimes difficult to sort out what actually constitutes an "invention" in some cases.

Re:What constitutes invention?

[Hooya]

> by marrying ... a stove and a fucking door.

If the stove didn't consent to the marriage it might be declared null and void. It might even constitute a criminal rape charge if the door had actually performed coitus as a "fucking door" is wont to do.

Didn't know Ben Franklin was kinky..

Re:What constitutes invention?

[cgenman]

The internet is (mostly) routed off of *nix boxes. Someone then created wireless routing. Wireless routing through Linux is pretty much a no-brainer at that point. Not because it's obvious, but because that's basically how you would do it. Wireless routing on Linux is about as revolutionary as someone invents the house made of wood, then someone patents the house made of wood built with a hammer.

Re:What constitutes invention?

[tagno25]

I have Prior Art for your patent. NASA, Hitler, TV, Radio, HAM, etc...

Re:What constitutes invention?

[Rennt]

Built? Sure. Assembled? Fine. Extended someting? Absolutely. Invented? I honestly don't know.

The word I would use is "engineered".

Actually, the "Engineering Test" would be a good way for the patent office to test the "non-obvious" requirements for patentability. You submit the function/requirements of your device/invention to a panel of engineers selected from your industry. If one of the engineers independently develops your "method", the method cannot be patented.

Re:What constitutes invention?

[butlerm]

If it was so easy, we wouldn't be reading the article and debating what exactly "invent" means in this case.

The only people who think that this is some sort of invention worthy of a government granted monopoly are either clueless or have a pecuniary interest in destroying the software industry. Ultimately, it is all Congress' fault of course, for being so naive as to establish and promote such a irreparably counterproductive system in the first place. Patents are inextricably evil precisely because the law provides no solid ground to reject trivial "inventions" like this. The only people that benefit in the long run from the practice of granting trivial patents are patent attorneys, leaching off the productive sector like the parasitic creatures of government malfeasance they are.

Used car sales people at least perform a public service. The patent bar, by and large, serves no purpose other than to cripple the progress of science and the useful arts, and make us all poorer, sicker, and less well off than we would be otherwise. Wake up in the morning and consider, how can I impair the economy the most efficiently today? File a patent application of course.

Obviousness?

[Fnkmaster]

Do they have to prove exact prior art, if they can prove that the differences between the prior art and the patent are obvious to a practitioner of the art?

A patent really should only cover non-obvious extensions of existing ideas. As the guy himself states, there was nothing in what he did that was particularly non-obvious to any person who did networking at the time, so it's silly that it should even be patentable, let alone an example of prior art. Rather, their effort just happened to occur at a point in time where wireless networking was just being introduced.

So why on earth should merely combining the words "embedded", "Linux", "wireless" and "router" make something patentable? These are not novel combinations that required a leap of creative insight, but rather extremely straightforward and obvious combinations.

Re:Obviousness?

[DeadboltX]

Exactly.

Imagine that chronologically white bread came first, and then sliced white bread, and then wheat bread. Well now someone has come along and patented sliced wheat bread. We already have wheat bread, and we already have sliced bread, isn't sliced wheat bread a logical next step?

We already had embedded linux devices, and linux wireless routers, why is combining these two things patentable?

Re:Obviousness?

[Grond]

Do they have to prove exact prior art, if they can prove that the differences between the prior art and the patent are obvious to a practitioner of the art?

Proving obviousness requires showing that all of the elements of the claimed invention exist in the prior art or can be shown to be within the common sense or common creativity of one having ordinary skill in the art at the time. As you might imagine, there's a lot of subjectivity to the 'common sense or common creativity' part, but there must still be some rational explanation for why elements that can't be explicitly shown in the prior art would have fallen under common sense or creativity. The bottom line is that evidence must be presented. It's not enough to just have someone stand up and say "it's obvious."

So why on earth should merely combining the words "embedded", "Linux", "wireless" and "router" make something patentable? These are not novel combinations that required a leap of creative insight, but rather extremely straightforward and obvious combinations.

In this case it appears to have been a failure of the PTO to find the relevant prior art. The burden is on the Patent Office to demonstrate that an invention claimed in an application is not patentable, and the Patent Office is often not the best at searching non-patent literature (i.e., everything that isn't a patent or patent application), especially when the NPL is also not part of a regularly published scientific journal. This leads to a lot of open source projects falling below the Patent Office radar. As you can imagine, this is particularly problematic for patents that were examined before decent search engines like Google became available.

For example, the bulk of the examination in this case occurred between October 2001 and November 2003. Google had just started to take off at that time (the IPO wasn't until August 2004).

Re:Obviousness?

[TooMuchToDo]

Is it me, or does the USPTO simply suck at finding prior art? I'm not saying it's their fault, I'm simply saying it's difficult for them to sift through huge amounts of data for each patent looking for the prior art that may invalidate said patent.

Re:Obviousness?

[Grond]

Is it me, or does the USPTO simply suck at finding prior art? I'm not saying it's their fault, I'm simply saying it's difficult for them to sift through huge amounts of data for each patent looking for the prior art that may invalidate said patent.

There are several reasons for this. One reason is that most examiners are inexperienced. Right now something like 80% of examiners have been working for the PTO for less than 3 years. Only about 7% have been working there for more than 10.

Another reason is that, particularly with regard to software and software-related inventions, the PTO has a hard time recruiting qualified examiners. A big reason for this is that examiners have to live in the Alexandria, VA area, which is about as far as you can get from the west coast IT industry.

I've also heard that examiners don't have the resources they need to do a proper search of the non-patent literature, but I don't have any hard data on that.

And, of course, sometimes patent applicants represent their invention one way during prosecution but then try to use it differently during litigation. For example, I believe Yahoo was sued or at least threatened for something to do with its maps website by a company that had a patent on using a computer to map...migratory birds. The claims appeared to read on Yahoo Maps, but it was quite apparent from the application as a whole that the invention was unrelated to anything Yahoo was doing. There's really not much the PTO can do about that.

Re:Obviousness?

[TooMuchToDo]

Possible solutions:

1) The USPTO needs satellite offices in tech hotbeds. San Francisco/Palo Alto, Austin, etc. Bring them to Alexandria for training, but have them work collaboratively remotely.

2) Contract with Google to build an app that manages the patent application workflow. Have it crawl for data that may indicate prior art, etc.

Re:Obviousness?

[h4rr4r]

I've also heard that examiners don't have the resources they need to do a proper search of the non-patent literature, but I don't have any hard data on that.

The patent office blocks google on their work computers?

Re:Obviousness?

[Grond]

The USPTO needs satellite offices in tech hotbeds. San Francisco/Palo Alto, Austin, etc. Bring them to Alexandria for training, but have them work collaboratively remotely.

Unfortunately it literally takes an act of Congress for the PTO to be able to do this.

The Telework Improvements Act of 2010 (aka the Telework Enhancement Act of 2010) would allow examiners to work from anywhere in the country. While the bill has passed the House and Senate it still has to go through conference committee and be signed into law. That might happen during the lame duck session or it might happen next spring. Or it might never happen, since it's not exactly the highest priority right now and politicians prefer to argue over stuff that gets headlines.

Re:Obviousness?

[TooMuchToDo]

House version: http://www.govtrack.us/congress/bill.xpd?bill=h111-1722

Senate version: http://www.govtrack.us/congress/bill.xpd?bill=s111-707

Grond: Thanks for the bill name!

Re:Obviousness?

[Cylix]

I was doing wireless routing prior to 98.

It just happened to be more expensive then it is now. The wireless devices merely formed the bridge link and then behind the link was an actually ipchains set of rules. It was quite a simple little nat box with some static routes to here and there. Embedded devices and linux were not a new concept then, but hardware support was certainly not as good as it is now.

Really the limitation of the day was having access to cheap commodity hardware and not access to the ideas. At the time I worked on a project to create an embedded distribution for a linux based weather recording and reporting system. The idea was to distribute the recording devices to several facilities to capture data from a broad range of terrain. The only reason we used ethernet was because there was a time when wireless cards were in excess of a 100$ and they were awful.

Now I feel like an old man.

This is good!

[countSudoku()]

Our submitter is doing the right thing here; testifying against a company that tried to patent an already freely available idea. No patent should be allowed to protect this device's software. The hardware is a separate issue. Thanks for doing the right thing!

Re:This is good!

[roc97007]

To which I would add, please let us know how it turns out.

Depends.

[blair1q]

It depends on what they're claiming.

I don't have time to read TFA or TFP, but I'm guessing that their implementation does not fit entirely within your list of suggestions.

If they're claiming everything, they were probably wrong in the first place, since your description would have relied on a lot of stuff that was public knowledge anyway.

But most patents are overbroad and depend on prior patents and common knowledge. It will probably come down to a detail that nobody has included before.

Re:Depends.

[blair1q]

Okay, now I've had time (I had to pee), and no, they're not the same thing. Some of the claims will succeed.

Though what's more interesting is that they call their software box "UNIX-derived".

Which may be a problem for Linus Torvalds...and a boon to SCO...

Re:Depends.

[Archangel Michael]

"UNIX-derived".

Is that marketing speak, or part of the actual patent?

If it was part of the patent, I would think that it would violate the whole of it. Linux is not Unix Derived. Because it is not Linux Derived, it has no bearing on Linux or SCO other than it was someone saying something that was factually incorrect.

Re:Depends.

[tepples]

Is that marketing speak, or part of the actual patent?

The former. I read the first claim, which the lawsuit reproduces in full. It does not limit the scope to operating systems that implement APIs listed in the Single UNIX Specification.

devide by zero error

[sgt scrub]

was this story designed to create an infinite loop in readers brains or what?

Re:devide by zero error

[Anonymous Coward]

Devide? It's spelled "divide." What the hell happened to English class in this country?

Re:devide by zero error

[AndrewNeo]

The same thing as math, apparently.

Re:devide by zero error

[blair1q]

Both were cut from the curriculum to make room for Creationism and Bible Study.

The Mother of all Patents ...

[PolygamousRanchKid]

"A Method and Process of using Linux to do Stuff with Things."

World domination follows . . .

Re:The Mother of all Patents ...

[stokessd]

> World domination follows . . .

Except on the desktop... :D

Re:The Mother of all Patents ...

[micheas]

> World domination follows . . .

Except on the desktop... :D

Embrace the server space, extend the server space to the cloud, extinguish the desktop.

Re:The Mother of all Patents ...

[Mister Whirly]

You mean like mainframes and terminals? Yeah, I think that has been tried before.

Re:The Mother of all Patents ...

[micheas]

Hmm, I didn't think anyone would take a reply to a +5 funny post seriously.

Fireplug Computers

[rcpitt]

Stuart Lynne and I were partners in Canada's first ISP, Wimsey.COM

After we sold that company - and the purchaser tanked, we started up Fireplug Computers Inc. doing Linux for embedded devices, including "Thin-Linux" which was specifically oriented to being a router. I ran a version of that on a 486 system until it finally died earlier this year.

We had wireless capabilities in this, with drivers for a couple of the then available wireless chips.

Fireplug was sold to Lineo Inc. in 1999 - and I'm fairly sure that Lineo did some work on Linux wireless too.

Who cares?

[RyuuzakiTetsuya]

I say we find everyone who submitted patent ideas before 2002 and buy them a drink.

Nothing's been more of a driver of IT support than those awful little boxes. Thanks for the job security guys!

How novel

[TopSpin]

There were 802.11 networks serving whole cities by the late '90s. All you needed was a high gain dish and line-of-sight. These guys were up and running in 1999. Every one of them had some form of *nix based "wireless router" setup. My own was FreeBSD, so clearly that has no relevance to a patent for something as brilliant as an integrated Linux wireless router.

Re:How novel

[gstoddart]

My own was FreeBSD, so clearly that has no relevance to a patent for something as brilliant as an integrated Linux wireless router.

Actually, in the technical drawings they claim it to be a "Router CPU with UNIX derivative operating system" -- so, your FreeBSD would have violated this patent if it didn't exist before the patent was filed.

I still continue to be baffled by patents. They invented none of "embedded", "wireless", linux" or "router". Doing it for the first time is cool (and mad props to the guys who were doing this and might bust this patent), but assembling well known components to do a well known job, but in a brand new configuration is an application of technology, not an invention.

Linux Router Project of Course

[Diesel Dave]

Certainly not the very first ever made, but likely the first mainsteam implementation that was available.

I had a 2U 386SX 16Mhz Workstation with full length ISA 900MHz WaveLAN card, that ran LRP off 3.5" 1.44MB.
Host name was 'Brain-Damage'. Some of the first LRP development was done on that back in 1997.

The boys over in Latvia that went on to form RouterBoard were doing much more then me with wireless but I'm not sure if it was with Linux at the time.

Dave

Re:Linux Router Project of Course

[b0bby]

Yeah, LRP was what came to my mind too. Ah, the good old days...

Re:Linux Router Project of Course

[NewWorldDan]

Well, you've got me beat. Best I can offer is an NT4 Server based router circa 1998. However, I don't think I added a wireless card to it until about 2001. And then ditched the whole setup for a retail router and wireless bridge shortly thereafter. Problem is, I didn't document any of this. I was just another college dropout at the time, and if I could do it, it certainly ought to qualify as obvious to "one skilled in the art." However, I don't think the USPTO has any sense of that phrase whatsoever.

I recall undertaking various networking projects aroundt that time to improve my own understanding, including work on Linux, BSD, and Solaris. Can't remember too much of what I did, but I did a lot of stuff.

Re:Linux Router Project of Course

[altinos.com]

When I worked at Ford in 1995-1996 we were playing with wireless connections and UPS backups to move diagnostic PCs around a repair bay without tripping on wires. These diagnostic computers were already pretty old, 386 PCs running OS/2. I don't remember what the wireless routers ran though.

Re:Linux Router Project of Course

[Diesel Dave]

LRP had the wireless drivers, a menu interface for routing configuration and was also embed-able, running off a read only boot medium.
The OP is asking for these specific things, and LRP had it out of the box, though I'd have to check exactly when the first wireless drivers made it in.

So LRP it was an existing distributed 'product', not a one-off box someone made. The former gives grounds to invalidate the patent.

Dave

Re:Linux Router Project of Course

[mmj638]

The title of the summary omitted the word 'embedded', present in the original article's title.

From the article,

I'm now nearly certain that ... we actually created the first recognizable "embedded Linux wireless router". PLEASE: Note the word choice, there - embedded, Linux, wireless, router. Eliminate any of those words and you end up with a different product, from a different person.

Combining words

[zooblethorpe]

So why on earth should merely combining the words "embedded", "Linux", "wireless" and "router" make something patentable?

Ah, but now that you have combined these words, you are infringing on my copyright to the phrase "embedded Linux wireless router"! I demand reparations for this flagrant misappropriation of my intellectual property!

(Sure, I'm being silly, but I'm also trying to point out the inherent ridiculousness of these attempts at "owning" ideas.)

Cheers,

What constitutes invention, anyway?

[MichaelKristopeit 12]

whatever the judge at the highest level court that will hear your case says it is.

Re:What constitutes invention, anyway?

[cm613]

If and only if you can afford to fight your way there. If not it is what the other guy, with the established business, says it is.

Re:What constitutes invention, anyway?

[cm613]

Thank you for your kind words. I can tell by the way your thoughts transform into free flowing words and by your condescending tone that you are most likely not an idiot like me and you have probably had an easy life.

I too am bitter for some of the things in this world. For example, patent law and corporate control of government policy, which are not mutually exclusive. I don't take it out on anyone I just throw out assertions on the way I see things on message boards such as this. I take it that you don't agree and that's okay with me, you are allowed your opinion.

I disagree with your argument about applying the law yourself. An independent inventor may not have the desire or the ability to apply law and spend time arguing in a courtroom. He or she may be more interested in building a business around something they invented or inventing other things. He or she may not be capable of decorum required in a court of law because they can't communicate without calling people names like idiot, ignorant or hypocrite. Established businesses, on the other hand, have legal departments that are paid to quash other businesses competing with them or else collect royalty from them. In this scenario, patent law is the instrument to make the rich richer and the keep the emergent down. That's the fundamental thing I don't like which lead me to make my comment.

I admit...

[AchilleTalon]

I did it.

Re:I admit...

[guruevi]

I did it earlier. I remember doing it when I was young (the early to mid-90's) with a WaveLAN card. If I remember correctly, it had only 1 Mbit/s throughput or so (our coaxed LAN was very fast at ~8 Mbit/s) and was very susceptible to the newfangled "magnetron oven" being used by my mother. The card actually did not have Windows 95 drivers so we had to add the drivers in the DOS environment (config.sys & autoexec.bat) and use DR-WebSpyder instead of Internet Exploder.

Am I missing something?

[PCM2]

The patent in question seems to be for something specific and somewhat different from a regular ol' Linux wireless router, like we all use. Particulars from the patent text (all emphasis mine):

• The device acts to provide connectivity between wireless backbone access points
• It is a primary objective of the present invention to provide a piece of wireless equipment that can effectively connect a large WAN
• Still another objective of the present invention is to provide wireless connections designed for outdoor use and flexible security

There are other particulars, but this seems more like something to provide large-scale outdoor wireless infrastructure than simple home routing. The diagrams also show little pictures of houses being served by multiple routers as described in the patent.

Re:Am I missing something?

[Bill_the_Engineer]

Still another objective of the present invention is to provide wireless connections designed for outdoor use and flexible security

Tell me more about these weather-proof radio waves and where I can purchase one.

Re:Am I missing something?

[Anonymous Coward]

If you are familiar with the HowTo referenced in the article you'll know that they set up their wireless router(s) to provide connectivity to a location 13 miles away from a DSL uplink to the internet. As the article mentions, the HowTo was actually hosted on the far side of that link. Whether it was technically set up as a bridge or not I don't know, but it provided connectivity in an *outdoor* setting between two wireless endpoints, at least one of which was connected to a very large WAN (the internet.)

The patent may have been filed with other uses in mind - and a patent is generally filed to cover as many uses and ideas as possible - but it is being claimed to apply to the wireless routers we all use, in large part because they use all the same technologies (and I used early Linksys and D-Link wireless routers with antennas in 2002 to provide similar links. I wasn't anywhere near the first to do it, but I read the author's HowTo many times because of it.)

Re:Am I missing something?

[wiredlogic]

You just quoted excerpts from the summary of the invention. It goes through such drivel as explaining the OSI model and other known-in-the-art fluff intended to bog down the examiner. That text has no bearing on the validity of the patent. The claims are all that matter, with claim 1 being the most important. That claim doesn't cover any new ground and just mentions an obvious combination of networking components, packet routing, and an authentication mechanism as provided by the 802.11 standard.

Re:Am I missing something?

[TheThiefMaster]

The 2.4GHz band used by wifi doesn't go through water (i.e. rain) very well. Therefore they could be described as not being weather-proof. I think the 5GHz band doesn't have this problem, and so could be described as being weatherproof.

What?

Stupid Patent Office

[morgauxo]

It's ridiculous that something like this could even be patented in the first place. Routers existed. Wireless existed. How obvious is a wireless router?

IEEE 802.11

[jklovanc]

How can someone get a patent on an obvious implementation of a new standard, in this case 802.11? I wonder if Anthony Spearman or Andrew Tompkins had any access or input into the standard? I wonder if they were part of the standards process but realized they couldn't patent the standard so the patented the implementation.

Claims in HTML

[bill_mcgonigle]

a TIFF? Seriously?

via Patent Storm:

Claims

What is claimed is:

1. A wireless provisioning device for use in public domain networks wherein the wireless provisioning device is accessible by a user of mobile computing devices, comprising:

a chassis;

at least one network card;

at least one wireless card;

at least one processor;

an operating system, the operating system operably configured in the chassis to control the at least one, network card, the at least one wireless card and the at least one processor, which are operatively coupled with the chassis;

a packet-switched interface capable of receiving a multiplicity of inbound framed packet-data to provide inbound packets and transmitting a multiplicity of outbound framed packet-data comprising outbound packets;

a channeling controller, coupled to the packet-switched interface that channels the inbound packets based on the inbound address information and constructs the outbound packets and channels the outbound packets with the outbound address information, the channeling controller capable of being effectively connected to at least one network via the operating system; and

an authenticator in operative communication with the operating system to allow authentication at the wireless provisioning device;

whereby the user of a mobile computing device connects to the wireless provisioning device without having to first access the Internet.

2. The wireless provisioning device of claim 1, wherein the channeling controller routes the outbound packets.

3. The wireless provisioning device of claim 2, wherein the channeling controller routes the outbound packets.

4. The wireless provisioning device of claim 1, wherein the channeling controller bridges the inbound packets.

5. The wireless provisioning device of claim 4, wherein the channeling controller bridges the outbound packets.

6. The wireless provisioning device of claim 1, wherein the operating system of the wireless provisioning device is an open source UNIX based system.

7. The wireless provisioning device of claim 1, wherein the wireless provisioning device further comprises a second processor.

8. The wireless provisioning device of claim 1, wherein the wireless provisioning device further comprises a memory device and a storage device.

9. The wireless provisioning device of claim 1, wherein the network card, the wireless cord, the processor, the operating system, the packet-switched interface, and the channel controller are operatively disposed within the chassis of the wireless provisioning device.

10. The wireless provision device of claim 9, wherein the authenticator is operatively disposed within the chassis of the wireless provisioning device.

11. The wireless provisioning device of claim 1, wherein bandwidth to individual user can be controlled by the wireless provisioning device operating system.

12. The wireless provisioning device of claim 1, wherein the protocol type of an individual user con be controlled by the wireless provisioning device operating system.

13. A wireless provisioning device, comprising:

a chassis;

at least one network card;

at least one wireless card;

at least one processor;

a LINUX operating system, the operating system operably configured in the chassis to control the at least one network card, the at least one wireless card and the at least one processor;

a packet-switched interface capable of receiving a multiplicity of inbound framed packet-data to provide inbound packets and transmitting a multiplicity of outbound framed packet-data comprising outbound packets;

a channeling controller, coupled to the packet-switched interface that channels the inbound packets based on the inbound address information and that constructs the outbound packets and channels the outbound packets with the outbound address information,

ReseauCitoyen.be

[Scotch42]

we did that at ReseauCitoyen.be several years ago. At that time there was nothing nor on the Internet nor elsewhere... We never claimed invention for that as this was obvious for us who where in the field... But we build some devices... We had some design published as public domain ones in hope to see them builded by corporations. And finally that was the case. So I can not understand that someone is now claiming a patent on such kind of ideas... PS: I'm not an english language native, so forgive some faults in my wording!

Re:ReseauCitoyen.be

[dasdrewid]

NO! BURN IN HELL FOR'NER!

Kidding! Don't worry about it, it's much better English than probably most of the people on here can do French. If you care, the only things incorrect are the two "nor"s in your first sentence and the way you conjugate "to build" ("built" being the past tense you want instead of "build" and "builded" as you have). Also, I guess you do need an "s" on the end of "design", as "some" requires a plural noun to go with it and "design" is still singular.

And I'm done being a nit-picking jerk... (I really don't mean any offense, just offering tips!)

Nice Job

[Anonymous Coward]

Just an FYI. You just tainted your testimony. Anyone reasonably worth his/her salt as a lawyer looks online to watch discussion of it by witnesses. If there is enough "chatter" your testimony can be thrown into a very poor light. Remember, if you are an expert, why are you asking a bunch of internet board posters what you need to tell a judge?

Re:Nice Job

[nickersonm]

He already gave his deposition in August. It seems that he is simply asking the question here because he is curious. IANAL and have no idea if that can affect the use of it in the future, though.

From the blog:

After giving my deposition, I've thought deeply about what happened in wireless and Linux from 1998 forward, and done a bit of independent research. I figure, maybe, by publishing what I know so far, more of the history and prior art behind the "embedding Linux in a wireless router" idea will come to light, and head off the second patent at the pass.

Also note that he's asking for different examples, not about the example that he is the source of.

Re:Nice Job

[camperdave]

Not necessarily. All he's asking is who was first. Once he knows that, he can search for specific documentation relevant to that fact. That documentation is what would be presented to the court/board/triumvirate/whatever at the patent office.

Re:Nice Job

[rtfa-troll]

I don't think he's an expert witness; he's a witness of fact. He doesn't need to be impartial. Consider for example:

I saw the bastard; he's the one who did it; I saw him put the knife in her back, I was two feet away and I'll never forget the evil grin in his face.

This is effective testimony, but has no level of impartiality. You don't impeach such a witness by saying he's biased. You have to show he's either lying or confused.

Re:Nice Job

[mtaht]

I am aware of the problems I may have introduced by speaking openly about the case, if it comes to a jury trial. However, I was engaged as a fact witness, not as an expert witness, and a deposition was also taken from the co-author of the howto. In that light, I felt it ok to ask the questions of my "tribe", publicly, that plague me, in the hope that I (and others) might learn from them. I am grateful to all the commenters here today that have taken time out to discuss the issues to the best of their knowledge and abilities. I have learnt more about patent law, prior art and about more prior art in the wireless world, in half a day, via slashdot, than in a month and a half of part-time research. There have been a bunch of links and other useful material posted here today that are GREAT. I will (ultimately) summarize on my blog, but quite possibly not until after the case is closed. Thanks again, all!

Who invented it?

[stacybro]

Wasn't it Al Gore?

How to define "Obviousness"?

[hackingbear]

All ideas, novel or not, are combination of prior ideas. And are certainly expressed in combination of existing words and shapes.

The fundamental challenge for the patent office, and those who are against patents, is that criteria like obviousness, similarity (need to prove assimilation to prior arts,) creativeness, etc. are fundamentally subjective. One could claim Einstein's Theory of Relativity is trivial and obvious, after hearing the details of the Theory; yet it was Einstein, and not billions of people before him in human history, that nailed down the Theory.

I don't think non-subjective definitions of these terms exist. At the end, these come down to who can argue better on a case by case basis.

Re:How to define "Obviousness"?

[Fnkmaster]

Of course those terms are subjective. But there is a very significant different between something like Relativity, that required (at the time) both a creative leap and connecting obscure mathematics with physical intuition, and combining several elements in a manner that is entirely within their design parameters.

In other words, I would agree that there are obviously many borderline cases that are questionable and need to be debated and argued. But I believe the Theory of Relativity unarguably required a creative leap. On I believe that taking existing router software that was already a core part of the Linux operating system, which already ran on embedded systems, and connecting it to a wireless networking card, which somebody had already written drivers for, contains no new *ideas* at all and therefore required no creative leap.

My point is that this isn't even the straightforward combination of existing ideas into a fundamentally new idea, which we could argue in terms of patentability - rather it is a straightforward combination of existing engineering elements into a system in essentially the way that the elements were intended to be used.

Re:How to define "Obviousness"?

[roju]

Sure, Einstein single-handedly invented relativity. That's why all the components are named after him, like Maxwell's equations, the Lorentz transform, Minkowski spaces, etc.

KA9Q NOS - circa 1985...

[Temkin]

The ampr.org domain dates to April 1988. Phil Karn's KA9Q NOS claims to date back to 1985. I know I established a routed connection from the east bay to Cupertino via a KA9Q "router" in San Jose using 1200 baud modems on 2m VHF radio around 1990 or 1991, and I was just repeating work that everyone else was doing.

Temkin

Re:KA9Q NOS - circa 1985...

[reiisi]

Yeah, just how different is a current router from the old ham data relays?

Re:KA9Q NOS - circa 1985...

[Temkin]

Yeah, just how different is a current router from the old ham data relays?

KA9Q NOS was a MS-DOS application running on a full size PC. It could route between a SLIP connection from the Internet to a HF/VHF/UHF radio "network", thought it was illegal to set up such a configuration. It was a reasonably full featured TCP/IP implementation, but there wasn't a lot of ability to add services. I think it had the ability to telnet out, and host FTP. It may have had naming services and possibly something like gopher, but that's not really needed to be prior art. What it was is TCP/IP routing on a wireless network. It suffered from retransmit problems on long haul links. Picture 3 stations in a line, 30 miles apart. The station in the middle hears both. The stations on the ends only hear the station in the middle. They implement CDMA and retransmit. The end stations step on each other and the middle station gets nothing but collisions. 802.11 came years later, and succeeded because it was a local area only network, and used unlicensed spectrum with no license requirement or legal restrictions on use.

So in 1988 we had wireless TCP/IP routing. The AX25 drivers in the Linux kernel were integrated into 2.0 circa 1996, and available as a patch some time before that. Jeff Tranter wrote an article for it in Linux Journal circa 1997. I believe the Linux filtering and routing stuff was pretty advanced at that point. So the Linux portion pre-dates the patent as well.

All that's left is "embedded Linux", But I'm sure there was someone running Linux on a precursor to a Soekris board back in the 90's.

What does that leave in the patent claims?

Everyone has their own definition

[jd]

I tend to prefer reserving the term "invention" for where the bulk of the work is totally new and "innovation" for where the bulk of the work is simply a refinement on something that already exists. (Under this definition, there is very little in the way of true invention, which is actually quite realistic. True inventors are extremely rare.)

However, these are not the legal definitions. In this situation, it is the legal definitions that matter, no matter how anyone else defines the terms or how rational or irrational the legal definitions may be. (Obviously, if my definitions were used, the patent system would not have moved past the four digit range.)

What constitutes invention, anyway?

[cm613]

In our economy: Invention is the right to a revenue stream if you can afford patent law litigation.

"Invented"? LOL

[Yfrwlf]

Wow. So whoever is first to put an operating system on a computer gets to patent that? That somehow that combination is magical and to be serious again for a second "no one in the industry would have thought that doing such a thing would be obvious"? No, wrong, fail. Die software patents. Just die. I don't care if they are in Linux's favor or not. All of them need to die, and most or all "normal" patents too.

ok, I'll spell it out

[ciaran_o_riordan]

Yep. He's looking for prior art that pre-dates him, and he's asking how to prove and attribute prior art to someone.

For this, there are some real common mistakes to avoid:
* You have to find prior art for the claims, not the summary
* You have to find prior art for *all* the claims
* Good news is that acceptable forms of prior art include ads, manuals, magazine articles...

It's not rocket science, but a lot of news stories like this end up wasting people's energy because people contribute their knowledge without knowing these simple rules, and it's all or mostly useless.

As someone who's spent time scraping slashdot stories with 200+ comments for possible *useful* prior art, I can tell you that a lot of people don't know these basic rules.

Re:ok, I'll spell it out

[Nerdfest]

Seriously, shouldn't you be able to invalidate a patent by showing prior art for only one claim? They should at least have to re-submit without that claim.

Re:Linus Torvalds

[MightyMartian]

I'm not even sure what this means. There were routers before Linux was even an apple in Linus's eye, running embedded operating systems. Why does the particular operating system even matter?

Read the F*ing claims!

[epa]

It is only the claims that matter - the rest is normally irrelevant to anyone!

It is quite a narrow patent that seems to be looking to provide a spoof internet connection to wireless devices when the can connect to the claimed device but there is no available Internet connection.

I have no idea how anyone could possibly infringe this patent, given that the last-but-two and last-but-one clauses of claim 1 are virtually unintelligible. The later independent claims are even stranger and more obscure.

IAAL!

(But as a patent attorney in Europe, I think that the USPTO and US patent law is strangely weird and broken.)

'cause its a WIRELESS router...

[mengel]

And in the minds of the Patent Office that's somehow Really Different...

Re:'cause its a WIRELESS router...

[MightyMartian]

Well, guys have been doing radio-based packet routing since at least the mid-90s (I knew a guy who worked on this for HAM radio). So that's a router hooked up to a radio pushing packets four or five years before someone dreamed up putting a wireless card inside a Linux box.

Access to non-patent literature...

[mengel]

A while back, I signed onto the peer-to-patent website for awhile, and tried to add some prior art references. I tried to refer the patent examiners to a Communications of the ACM article from 20 years ago, and they said they didn't have access to that and I'd have to get them a PDF(!) Similarly they couldn't seem to come up with a copy of Karrels & McKusic et al. to see what was in 4.3BSD a quarter century ago. I mean, they ought to have a library, right? As a public service, I got an ACM membership again for a year so I could pull down the ACM article and give them a copy...

How are they going to recognize a rehash of old ideas if they don't even have the basic reference materials?

Isn't Linux in its self proof of prior art?

[edelbrp]

Unless the person applying for the patent wrote in the additions to Linux to begin with, wouldn't the fact that the functionality exists be proof of prior art? In other words, Linux has features included for the purpose of being used. Even if the features were used in an unusual combination, the fact that the features are able to be used in such a combination meant that the authors intended for there to be such flexibility for that to happen, and therefore even if not physically done before, at least the general idea and concept had been thought of and implemented in code... therefore prior art.

I understand there's probably more to the patent, such as hardware design and such, but I just find it interesting that somebody could patent, even in part, the functionality in Linux.

But does it answer the claims?

[rickb928]

Just remembering and poking around, I found these guys that were running wireless Linux routers in October 98.

The Open-WRT project was soliciting ideas at least in 1999.

Someone at HP was working on this in 1996. I don't know how far they got before 1998.

I'm suspecting there were other projects back then making good progress. But I can't tell how applicable these are. The HP project is sure interesting.

an example

[ciaran_o_riordan]

You can invalidate one claim (as it was written) by submitting prior art about that claim.

"as it was written" because the patent holder gets the chance to reword that claim and resubmit.

Invalidating one claim can solve your problems, but patents are written like thickets, so there are often many claims that cover a real world product.

1. A big box
2. The machine from claim #1, with a tap
3. The machine from claim #1, with a basin
4. The machine from claim #2, connected to tubing
5. The machine from claim #4, where the tubing is waterproof
6. ...

When you build a house or a petrol pump and someone says you infringe a patent, it's usually more than one claim.

Partial invalidation is possible, and might be enough to solve your problem, but it's a broad task. It's not the "Hey, this looks like the thing I saw in 2003" task that many people think it is.

Re:an example

[Anonymous Coward]

This is something that bothers me about patents, especially software patents. They don't patent one device, but a huge set of potential devices. I've see software patents that patent on the order of 10^100 unique solutions to a problem.

Linux on a toaster?

[talsemgeest]

I wonder if anyone has patented running Linux on a toaster...

Mikrotik was doing this years before

[Dryanta]

www.routerboard.com, www.mikrotik.com. Their RouterOS is very similar to what other companies do for specific applications. They were founded in 95 and produced their first hardware products in 02. http://www.routerboard.com/about.html

patent office shouldn't be so far behind the curve

[reiisi]

I thought about this problem a lot about 25, 30 years ago. How do patent examiners recognize innovation unless they already know pretty much everything under current development in each particular field?

I think some ambulance-chaser-turned-congress-critter must have been thinking about the problem and recognized the implications back then, got himself on the right committee, and started deliberately underfunding the PTO and doing other things to keep them from bringing in competent examiners. I do recall reading about examiners complaining that they couldn't get enough other qualified examiners, and I'm pretty sure I recall reading about one who says they more or less decided that working the backlog down was more important that making sure the patents were valid.

I also spent a bit of time thinking about the database that the patent office should have been building. Too bad I spent more time finding problems than solving them, or even than talking about them.

Re:Al Gore

[billstewart]

Al Gore wanted the National Information Superhighway to have 65 bits. That can handle a lot of trucks.

f.ck you both, patents must die

[anton_kg]

I have been using Linux as router since 1996 because routing/firewalling was the only one thing Linux was good for Now you are saying that I have to respect your patent to use my FREE OS? Seriously, go to hell. My old i386 Intel-based router was even more embedded then any other modern wifi router.

http://en.swpat.org/wiki/US7035281

[ciaran_o_riordan]

I was just reading the patent. It's a strange one. I can't see any difference between claims #2 and #3.

It seems to be a software patent. The hardware elements claimed are all non-innovative.

I've started a wiki page for it here. Not sure if this will turn into an article about the court case or about the patent, but this case seems to provide interesting examples of a few topics.

http://en.swpat.org/wiki/US7035281

(I'm working on it now, but have to shutdown very soon. Should get to flesh it out tomorrow.)

Embeded Linux router with wireless capability

[Lightning_lew]

FREESCO could possibly pre date the patent dates and it does meet the basic criteria. It is also still in production and still running on a floppy.