Slashdot Mirror
Bryar Takes On Patents And Their Friends
Jack Bryar's column over at Andover News comes out swinging; not at software patent holders themselves, but at the convoluted, corruptable machine of the USPTO. Bryar points out, among other things, that "the time available to Patent Office employees to process, review and approve a patent application -- an application which may run to hundreds of pages and be highly technical in nature -- has been reduced to less than eight man-hours." (Raise your hand if you think that's adequate.) Interesting, and mostly unflattering, information, too, about Patent Office head Todd Dickenson, and the changes which he's ushered in, or ignored.
You can use this script instead.
----
Stop worrying about the risks of nuclear power and start worrying about the risks of not using nuclear power.
After almost a year of being owned by Andover, Slashdot finally links to a story on Andover News, one of the most underrated tech news sites on the web. I've read fascinating articles (and some duds along the way, too, of course) there for the last several years, yet few (if any) of them have been linked here. Granted, there's the apprehension associated with giving the appearance of being "taken over by Andover" or "pandering to your own ad sales department" with putting too many links up, but really -- most of the columns written by Andover's "three Bs" (Bryar, Bresnick, and Blankenhorn) are substantially better than what usually runs in the "Features" section here otherwise :) Yet they typically get seen by a much, much smaller audience....
In a more "on-topic" light, I'm glad that Bryar had the guts to point out that the number one thing wrong with the patent system today is that the inmates have effectively taken over the asylum; it makes you long for the days when the nation's entire body of patent examiners were three members of the Cabinet who met a couple times a year. If I were to propose a fix for what's broken with the patent system, it would go something like: (1) Only legitimate innovation is patentable. Patent inspectors have to either deny a patent, or produce (and attach to the patent) a written description of why the patent is valid. and (2) Patents are valid for exactly TWO product lifecycles within a particular industry. Thus, software patents might last three years, while the term for pharmaceutical patents might be extended out to seventy. Yes, I know this second part is basically what Jeff Bezos said in his reply to Tim O'Reilly -- he was right. You can't abolish software patents entirely, there really is legitimate innovation (new techniques in wavelet compression, anyone?) in software that is and should remain patentable.
This is my opinion and my opinion only. Incidentally, IANAL.
MOO;IANAL.
There used to be a picture linked here.
I was a patent examiner, and hated it. Yeah it is great to see the new tech that is in dev, but there s no time to savor anything. I should mention that the amount of time that you have to work on your applications goes down as you go up in grade and get promoted. When I was a GS7 I had about 16 hours per application, and when I got promoted to GS9 I had about 12 to 14. I know that some people there were happy about remaining a particular grade as that was all they could handle.
The USPTO is one of the few goverment agencies that has quotas, that are that strict. They are also one of the few that actually make a profit, like the post office. Your taxes do not pay for a patent examiners income, it is payed through the patent application fees and patent maintenance fees. It can cost one thousands to get a patent.
I personally think that the only solution is to privitize the USPTO and make it gov regulated. One of the current problems they have is that they do not have enough examiners and Billy Clinton boy decidede that he would take money from teh USPTO to help balance the budget. Several millions of dollars that could have gone back into the USPTO to better train some of there examiners.
One last thing. Some of the examiners do a search if they find something great if they do not they allow it and figure that they can get soreted out in court. It is a screwed up system that desperately needs repair.
send flames > /dev/null
Only 'flamers' flame!
Try here http://www.patents.ibm.com/ibm.html This is the IBM intellectual property network
-Master Switch, one more element in the machine
The problem is not that the USPTO (or any patent organization) gives too little time to review software patents. The problem is the choice by the USPTO to allow software patents in the first place.
Bruce Lehman.
Lehman was the USPTO head at the time, took some heat for the decision, and even held public hearings (one on each US coast) to 'discuss' whether software patents should be allowed.
What a sham.
By the time I arrived to present at the west coast hearing (one day affair), it was clear that Lehman had already made his decision to allow them (even though he supposedly was there to hear both sides of the issue). What a considerate and professional fella.
Oh, he and his fellow lawyers.
There were no non-lawyers on the panel. The audience was roughly half lawyers (left hand side of the room) and software professional (right hand side of the room). That was about the break down of the audience.
The lawyers smelled deep pockets of money (Lehman included), career mother lode as it were.
The software professionals wanted the freedom to write and think unfettered and unencumbered by software patents (and any system of governmentally institutionalized artificial monopoly). Software is about writing and thinking. If you can't patent writing (that's what copyright is for) and you certainly shouldn't be able to patent thinking (right? though that's what software patents essentially do!), then you certainly shouldn't be able to patent software.
Nevertheless, Lehman decided in favor of the software industry subsidizing the legal profession involuntarily through software patents.
I remember some of the convoluted and conflicting remarks Lehman made while I was waiting to present (the USPTO in their consummate professionalism and organizational prowess had forgotten to schedule my presentation, so I went last). At one point Lehman suggested that the USPTO would not be weakening the software industry by taking the best and the brightest but only those from about the 90th (if I recall correctly) percentile. Therefore, according to him, the intellectual cost to the software industry would be negligible.
No kidding! Why would anyone with half a creative iota of software design work for the USPTO?
On the other hand, Lehman claimed that the USPTO would certainly understand what to patent and what not to patent as unique. However, when queried about how to check patent validity, Lehman would have no answer (unsolvable, unautomatable problem).
Everything about software patents was and is a catch-22.
The (perhaps ill-conceived) notion of patents was to provide a method for an inventor to maintain a property of his idea until finding a means to produce the idea for profit. Monopolies are illegal, but patents were supposed to provide incentive enough for innovation to outweigh the evils inherent in monopolies (that was the idea anyway).
Where's the difficulty in manufacturing (copying) software?
There is no difficulty. Software does not even approach the model where inventors don't have the means to produce their product ideas (remember, patents were invented in the 18th century, before the industrial revolution, so manufacturing capability was scarce). If you can copy a diskette (or download off the internet), you know how to manufacture software on your computer.
So, just the idea of software patents is wrong.
Lehman was told so. His lawyer panel was told so. His half lawyer (cheerleaders) were told so. He decide to bolster his legal industry's "customer base" (think: tax base) anyhow. Lehman was either phenomenally unclear on the concept (which I prefer to think) or very keen on expanding his profession's market base (which the cynic in me suspects was the case even if Lehman wasn't bright enough to know what he was doing).
The real answer should be to purge software patents from the legal and software industies.
-=-
To be clear, Lehman was in charge of the USPTO (US Patent and *Trademark* Office) at the very time when trademark domain names (i.e., squatter properties) were all the rage. Now that's an issue that the USPTO could and should have responded to immediately and with prejudice.
Instead, Lehman did nothing on the obvious (trademark violations) but created a whole new morass of patent law (software patents).
Congress did not create software patents. Lehman created software patents by rubber stamping the idea and setting precedent. He didn't do his job on trademarks (the USPTO didn't address this issue until nearly a decade later, I believe) but went out of his way to create an infinitely growing bureacracy.
The bottom line: Bruce Lehman may be bad, the new guy may be worse, but the person who hired/appointed them truly lacked leadership.
My nagging question is how does one become head of the USPTO (appointed? elected? random lottery from ladder climbing lawyers?)? The person hiring these people is unclear on the concept of making society healthy and productive. They have hired incredibly pig-headed and self-serving bureacrats. Lehman may be bad, but the person who hired him is truly a culprit as well.
The blind leading the blind.
I wish I had more good things to say about software patents, but I don't. Once they became official with Lehman's very public and official rubber stamp, they have and will only become worse.
Caveat emptor.
The sooner they're purged, the better.
= Joe =
http://www.andovernews.com/cgi-bin/news_column.pl? 533
Secondly, I would suggest that every US voter sends the above URL to every politician that he votes for, with a suggestion that his rep read the story. End your polite message with, "I shall be watching your response to this situation with some interest."
Votes are what count to an office holder. It's the best way for an individual to influence a politician.
For those of you who missed it, the NY Times Magazine on Sunday (march 12th 2000), had an article very critical of the handling of software patents. It talked about how the process was designed for a much slower, less litigous, and more tangeble technological world, and how the inspecters are not allowed to consult anthing other than the patent database when looking for prior art. Worth the read, if anybody has a URL to the online version, please do post it.
---
Play Six Pack Man. I
It has never been the case that an average examination was less than seven or eight hours. Never. For an examination fee of a few hundred bucks, the most that can be hoped for is what is called in the trade a "novelty" or "patentability" search (as opposed to a validity search). In most art areas, this is reasonably adequate.
A validity search typically requires far more resources, on the order of $10K-$50K or more. In litigation scenarios, a six figure search bill is not suprisingly uncommon. Since the corpus of prior art is virtually infinite in scope, one can spend as much as you want and still not have considered all the relevant art.
Congress made a determination to balance the quality of a patent search against the cost of access to the patent system. And for most art areas, this balancing is adequate.
Now as to the rhetoric. While an average eight hours are spent on examination, this does not mean that eight hours only will be spent on a multi-hundred page document. The number of huge applications are a small fraction, and even for large applications (at least in the software arts), they are large because of the multiplicity of claims (most of which are similar, and the patterns to which are obvious) filed as a result of recent case law. His remark that they are highly technical ignores that examiners only review applications in their particular area of expertise.
In short, it is perhaps most polite simply to note that Bryar was using his statistics improvidently.
Now, that being said, I for one acknowledge and agree that a novelty search is inadequate for certain art areas, in particular the software arts and methods of doing business. Too many patents are issuing when the best art had not been considered.
The real harm from this is that once the stamp is impressed on the deed, the claims are cloaked in a virtually impossible to overcome presumption of validity, even when killer art is available. The jury is instructed that unless "clear and convincing evidence" of invalidity is offered, they must find for the plaintiff. In practice, juries always find some doubt, and find the patent valid.
This is wrong and unjust, particularly when a defendant raised art that: (i) was not considered by the examiner; and (ii) which raises a substantial new question of patentability.
I have a proposal for legislation presently being considered in various fora, which I think may adequately strike more fairly the balance between keeping the scope and costs of examination in check, while protecting the marketplace. For prior art that fits the preceding two criteria, reduce the standard of evidence to that of a mere preponderance of the evidence. If the art is new, let the court consider it without enhanced evidentiary requirements, and thus let the plaintiff go to court at his or her own risk.
For more detail on the proposed reforms, check out this memorandum.
I'd have thought this was obvious. Do you really need it spelled out for you? Every time someone files a patent they are in essence telling you, "This thing you may not build without our consent." Sometimes that is appropriate, if, for instance, the patent holder spent a lot of time and money inventing the thing being patented, but more and more frequently the thing being patented is either bleeding obvious, or else so broad that it covers not only the patent holder's invention, but also a whole bunch of other things that he didn't invent. All of this translates to shutting you (yes, you) out of the market. (Isn't it amusing how everybody flaps their lips about how wonderful the "free market" is; yet, nobody actually wants to compete in it?)
Let's use one of your own examples:
But of course your solution is complete baloney because if your analog processor were so superior you would have invented it without having been forced into it by someone else's patent. For a company to patent their own revolutionary CPU design is reasonable; a patent that covers any CPU that someone else may invent subsequently is way out of line. And for a slightly sillier example you say:
But, what if somebody patents "the use of man made utensils to facilitate ingestion of foodstuffs"? Will you then just eat with your hands?
Burying our heads in the sand is not the answer. If we "stop worrying" and "just hack code", we could well find a cadre of lawyers at our doors telling us that we no longer have the right to hack code unless we pay up to some bozo who has gone and patented whatever we happen to work on. If that happens, far from "dying of its own excesses", the system will perpetuate itself as large companies that can afford to patent everything under the sun enjoy a legalized monopoly on writing software.
I think you fail to get the point - innovation is a path to find the BEST way to do something, not just the latest and greatest. If the "best" way suddenly becomes the private property of one commercial company, then even if you find something almost as good, you are still at a disadvantage. Amazon's "one click" patent isn't just for "one click" - they claim that ANY shopping cart method that stores the customer's details so the user can just commit their order, regardless of if they click a button, ring a bell or use voice recognition to shout "make it so!" from the other side of the room.
Overturning a patent once granted is a slow, expensive process with courts automagically giving patent-owners restraining orders on request that could completely destroy your income just when you need it to PAY the lawyers you have to hire...
What it comes down to is that patents, particularly software patents, are being given out by people with insufficient time to evaluate them, even if they HAD the skills to do so, which they don't.
--
-=DaveHowe=-
I think you're missing the point. This is the first article, out of hundreds I've seen recently on the patent problem, that actually addresses the "real problem", which IS the PTO. You can't blame corporations for greed, or for taking advantage of whatever corruption they can find. One can even argue that they're required to do so by law - officers of publicly traded corporations must operate with the best interests of their stockholders in mind.
Try this thought experiment: if no law enforcement agency in the country arrested or prosecuted anyone for murder, would you say that "the real problem is all these murderers"?? Or would you say the real problem is that the government is not doing its job?? It would be the latter of course, and it's the same with patents.
It's ironic that the patent system has become so corrupted that it has exactly the opposite of its intended effect. Not only is the inventor no longer protected by the system, she has actually become a victim of it, by being forced to defend her true innovation from the feeding frenzy of IP lawyers and their spurious patents. Open source inventors are especially vulnerable, because participating in the patent process in any way will cost you six or seven figures, which open source developers rarely have. This is why Linux IPOs are a good thing - so companies like Red Hat and VA Linux can hire patent attorneys to defend open source when necessary. OTOH, they also present a target for patent sharks to attack. One of the best defenses against a patent lawsuit is not having any money!
Of course, the real inventors these days have all signed over their entire brains to the companies they work for... but once in a rare while, they are actually rewarded for their effort. Usually not though, especially in Silicon Valley, where the vulture capitalists and the lawyers have perfected the art of the screw. The SF Chronicle ran a pretty good series on that topic recently.
Please explain to me why we're getting so worked up about it. Open Source / Free Software does NOT need patents. We design a program/device and release it to the public. If somebody goes off and patents it after that point, we just point to our reference model and say 2 magic words: "prior art", and the problem disappears.
We wish this were true. But it's not. Prior art can be used to demonstrate that a patent application is not valid. To overturn an existing patent on the basis of prior art will often require time in court - simply appealing to the USPTO is not enough. How many open source developers are prepared to spend lots of money fighting off patents which have been erroneously granted? Even the EFF has limited funds - don't think that we can always leave this to the EFF to cover us.
Patent reform, and a USPTO which doesn't view numbers of patents granted as a basis for productivity payments to the patent officials, is needed, badly. Just because you believe that the wolf howling outside your window won't eat you doesn't mean you shouldn't try and shut the doors.
Cheers,
Toby Haynes
Anything I post is strictly my own thoughts and doesn't necessarily have anything to do with the opinions of IBM.
I have two friends that work at the USPTO (not in the software division) and I am disturbed by how their compensation and advancement structure is so quantity oriented. Basically, they are told, you must review X number of patents in a week, if you do 110% of X, you get a raise of Y%, if between 100% and 110%, then Z% if less then just 3%. My friends are very competent people, but I think that the emphasis should not be on number of patents (or rejections) turned out but on thoroughness of patent review. Of course pressures from concerned companies would influence them, but I think cutting edge industries (such as software, biotech, etc.) will suffer as individuals begin suing each other needlessly. On the other hand, it's a great big universe and I'm just one tiny speck...
Y'know you take a perfectly valid argument like this, sprinkle it with some overenthusiastic "journalism" and then soundbite it into "a patent in the hundreds of pages gets an 8 hour review" and it makes it hard for anyone to take it seriously.
I totally agree with this argument. The USPTO is overloaded and has made some high profile errors lately. One piece of good news, is that they are using technology as well as they can to streamline the patent review process. But their is no way around good old common sense being applied to these applications and that takes time. Agreed.
However, soundbites like the one in the teaser for this article, can really misstate the situation. An application may run to hundreds of pages. Maybe most are only 10! Does anyone know the average? Patent Office employees get 8 hours to process, review, and approve a patent application. I bet you'd get really good at it after the first few thousand. I've seen bio papers peer reviewed in about that time.
Anyway...most regular Slashdot readers seem to have enough critical thinking skills to ignore this kind of slanted journalism and make their own conclusions. That's what makes our moderation and rating system work. Hey...maybe you could patent it and license it back to the USPTO for patent review.
BRENT ROCKWOOD, EST'd 1975