Slashdot Mirror
James Gleick On Software Patents
haase writes: "James Gleick has written a thoughtful and compelling piece on software patents for the New York Times magazine. This would be a good piece to send to your representatives. You can read it at
the NY Times Web site. (Registration required.)
"
A famous author can be full of shit as easily as a nobody!
Fight Spammers!
The About.com site here.
The sidebar is very good, but you have to hunt for the links if you are using a CSS enabled browser and the article as a whole suffers from formatting problems. Still it is a very good article and I really liked the illustration "Procedure for Simultaneously Walking and Chewing Gum" by Dugald Stermer.
Nonetheless there really isn't much new here for us Slashdot folks. That is, other than some really good new ammunition for the next time you want to talk about stupid patents. (I really cannot believe someone got a patent for measuring breast sizes with a measuring tape!) For us Mr Gleick is preaching to the choir.
Our real hope is that everyone else wakes up and realizes the danger stupid patents (and, perhaps, software patents in general) represent to our currently flourishing 'New Economy'...
Jack
- -
Are you an SF Fan? Are you a Tru-Fan?
I'm constantly worried now in my software projects that I'll need to waste all of the money on a patent to keep companies like Amazon from illegally patenting later what I already invented and shutting me out from use of my own invention. Prior art on Amazon's one-click shopping? Yes, it's called a vending machine, and it's been around for a while. I suppose vending machines are now going to have to add a button for "are you sure?" to keep Amazon away.
James Gleick is an outstanding non-fiction science author. He wrote Chaos, a very good introduction to Chaos theory and Genius, a biography of Richard Feynman (now don't say who the #$^*@ is Feynman!). I find his writing to be technical enough to be interesting, but not over my head. While not exactly written for the layperson, his books make science more approachable to us non-PhD's.
If anybody knows of other books by him, please let me know. I'll read anything he's written.
Admit nothing, deny everything and make counter-accusations.
...and check out the article at James Gleick's personal site, around.com:
http://www.around.com/patent.html
The old world is not going to go down without a struggle, nor should they be expected to. And so it's war, and they are coming after the internet, and all of its "free" this, and "open" that, with everything they've got. What this amounts to is an attempt to "fence in the frontier", and it is a direct and strategic attack on open source - you can't GPL it if someone's already patented it - and more broadly an attempt to kill off the ecosystem that has allowed it to flourish.
Someone needs to start the Gnu "prior art database" to catalog all of the unpatented implementations and algorithms. And sooner or later, the geek tribe needs to become a political force and kick some butt in Washington.
"Fig.1 - a device for knocking patent officer's heads together..."
I'm serious here... if some non trivial percentage of patent examiners were Open Source programmers, there would be two immediate benefits:
1) Stupid patents would be caught earlier.
2) We would have more time to write code.
Both are important here, remember what Albert Einstein did for a living while working on some of his early theories? It's basically a 9-5, gov't job with full benefits, and you have the rest of the time to yourself. No 80 hour weeks of unpaid overtime, leaving you too exhausted to do a little kernel hacking when you get home.
I think its time we got off our collective asses, and really did something to make a difference here.
Me.
http://www.around.com/
Admit nothing, deny everything and make counter-accusations.
Firing up a Google search will probably help. Whatever.
--
Industrial space for lease in Flatlandia.
But this is not longer true. It appears that the European Patent Office does not issue patents not because they believe that "software is inherently unpatentable", but because originally they believed that they did not have the equipment of skills to judge their quality!
IBM (all hail the mighty patent machine) lodged a software patent with the EPO last year and had it turned down (as per their plan one would assume) and then lodged an objection and while the patent was still refused the appeal ruled that the exclusion of software from patents is not valid under all circumstances
Which all means that companies are falling over themselves to get software patents out in the EU as well, left the good times roll
EPO Appeal search engine, search on software
IBM Appeal (PDF)
C.
I sometimes write stuff
James Gleick has also written "Faster: The Acceleration of Just About Everything". However, I thought this book was rather poorly written compared to his other two.
pooptruck
[a very well written article, IMHO] is:
Meanwhile, the dollars-and-cents reality of running the American patent office has also encouraged the patent explosion. In 1991, the patent office was cut off from general tax revenues and required to subsist entirely on fees for its operating budget. The political argument was that customers should pay for government services. Thus, officials think of their fee-paying patent applicants as their customers: the more the better, again. Examiners know that their year-end bonuses depend on productivity. Each morning, as Commissioner Dickinson arrives at his Crystal City office, he walks past a framed poster bearing the motto "Our Patent Mission: To Help Our Customers Get Patents."
It's virtually forgotten that government's customers also include the rest of the nation, the citizenry at large, whose fortunes depend on the agency's judgments and policies.
Also, to the guy spouting on about "Who the hell is James Gleick" - I think you've missed the point. It's common courtesy to attribute an article to an author (in the same way as software to an author, a movie to a director, a song to a singer, a slashdot post to the poster...)
Nobody said "by the wonderful James Gleick", or even "by the authoritative James Gleick".... work it out.
My 2 cent's worth.
--- cut: Eat well, exercise, die anyway.
at least not according to a lookup here altho I'd love to find a silly one like that in the database.
All I can add is, after grad. in '82 I went to the PTO and worked in the "info storage and retrival" section for a summer - it's a ruff job, slogging thru all the legalese and trying to shoot 'em down, but that's the examiners job. I had a few pat. applications, several actually, which were just 'burn a program into a 2716 ROM and patent the ROM' (REJECT!!). I actually was about to issue a Patent on a few but my supervisors said, "What's so new about this" and quickly produced a document that preceeded it if you interpret it broadly enough. They kept emphasizing 'broad' thinking - that is, if someone tries to patent a memory scheme that is implemented electrically in Si chips, you can reject it with a 'similar' memory scheme implemented mechanically in wooden disks and dowell rods.
try { do() || do_not(); } catch (JediException err) { yoda(err); }
This is a great article to help the average person figure out what's going on with software patents and such. Most of the things in this article have been discussed here on Slashdot already, but the article offers a lot of great insight for those who may have missed it.
:)
The ignorance about software at the USPTO is a huge problem. Possibly worse is the fact that the patent examiners have a lot of incentive to grant as many of these patents as they can. JG claims that examiners' bonuses are based on the number of patents they grant.
I'm not sure what the answer to this problem is, but there's no doubt that this is only going to get worse until something is done about it. It seems obvious that software patents are doing a lot more to stifle innovation than to foster innovation. It seems obvious that the big guys will reap huge rewards due to their ability to abuse the patent system so easily. As you can tell from the article, Commissioner Dickenson from the USPTO seems blissfully unaware of the problems with the software patents they are granting. In fact he seems downright pleased with the current state of affairs. Keep in mind that the USPTO's budget comes from examaning patents now--all of it if I understand correctly. And since software patents have become so popular the USPTO has been granting patents in record numbers.
In other words, don't count on the USPTO to try to change things for the better--unless they can find a way to do it that will allow them to issue more patents.
That's my read on it at least. Great article at any rate. The author even used the word "hacker" in a non-derogatory context. I'm impressed
numb
One point in the articles was very revealing to me - the pay and bonuses for the patent examiners comes from approved patent applications. This is a serious management mistake. A better approach would be to charge a signinficant fee for an application, refundable on approval but forfit if the application is denied. I doubt it would ever happen but it does strike me as fairer and more efficient.
Fairer because approved applications are presumably a public boon, so it is only fair that they be publicly funded, while denied patents are just a public nuisance consuming public officials time and energy, so it would only be fair to make the applicant pay.
More efficient because they would give the examiners an incentive to deny frivolous applications and fees could be set at a level that would be a disincentive to frivolous applications.
As I said I doubt it will happen, but it appeals to me because it would reverse a system that currently seems to be travelling flat out in the wrong direction.
Aren't what some of the patents cover actually coverable by trademarks? I mean, a series of shortcut keys for menu items (Brief/Star/Borland) isn't actually a /process/, but it is, in a sense, a mark of their product. Likewise, Apple's blue gel encasing is a mark of their trade, not a /process/ and not patentable (unless there is really something innovative in that casing as they would have us believe).
Where to trademarks play into this? I mean there has to be more use of tradmarks than to simply protect the word "Coca~Cola" (tm).
It's 10 PM. Do you know if you're un-American?
As the article mentions, Gleick's most recent book is /Faster: The Acceleration of Just About Everything/, and it is an outstanding book. I highly recommend it.
:-)
It even quotes from the Jargon File.
You cannot apply a technological solution to a sociological problem. (Edwards' Law)
The biggest problem I can see with software patents vs. open source is that open source has no revenue stream to use to finance a defensive patent portfolio. Though that leads to an interesting idea upon which others have speculated before: the Open-Source Patent Portfolio. Since one of the ways that companies (like IBM!) derive revenue is by licensing patents, the OSPP could patent inventions by open-source inventors (or have the patents sold to them), and license them out under a license which gives anyone the right to use the inventions in open-source software for no fee. Companies wishing to use any OSPP patents in closed-source products could do so for a fee and a cross-licensing agreement which allows all their patents to be used in open-source software for free under the OSPP license. This scheme might even generate enough revenue to kick some back to the inventors.
--
Time is Nature's way of keeping everything from happening at once... the bitch.
I know I'm following up to my own post, but I've just been looking around Gleick's homepage for the first time, and discovered that /Faster/ has its own domain:
http://fasterbook.com/
You cannot apply a technological solution to a sociological problem. (Edwards' Law)
I haven't read Chaos yet, though it's on my list, but Genius was one of the best books in its genre that I've ever picked up. I read it in high school, and much was over my head, but another round through it after four years of physics and engineering really amazes me. Well put together, interesting organiztion, and just well written... I just hopped over to www.around.com... turns out he won one of those funny little prizes (Pulitzer) for both of these. Guess we're not the only ones who think he did a good job 8^)
"It's tough to be bilingual when you get hit in the head."
Current practice, however, is to patent the "use of a calculating device and communications network to distribute news in a timely manner". Suddenly Slashdot, the NY Times, and dozens of sites owe me 0.25% (I wish), even though each uses different methods to implement the concept.
It used to be that people patented specific, and presumably better, designs for implementing a carburator. Now they patent the very idea of a carburator.
Aficionados of silly patents will rejoice in finding this patent, kneepad and all, in the patent database, just with a slightly different number. It's US Patent 5,993,336.
Also, for those who have browser troubles on the around.com version of the story, there's a stripped down version linked from Advogato.
LILO boot: linux init=/usr/bin/emacs
If the US Patent Office violates its constitutional mandate, shouldn't we be able to sue them? (Note that the constitutional mandate may be completely different from the statutory mandate - they may be following the laws passed by Congress to a "t", yet still violating the constitution). Constitutionally, the purpose of patents is "to Promote the Progress of Science and useful Arts". If the Patent Office is not doing this, then I would suggest that they are engaging in some kind of illegal restraint of trade.
BTW, I don't see any kind of change occurring until the megacorps decide that it is in their best interests.
A lot of people have been talking about doing something about broken software patent law. I think now it is finally time that we should actually do things. There are at least two things we should do:
1-- Establish a pool of open software patents. Anyone can use any patent in the pool providing they agree to make all of their own patents (if any) available to everyone else in the pool. proprietary developers who want to use some of the patents in the pool may do so: if they add their own patents to the pool.
2-- Establish OSS patent language: you may use this software, but any patented material you add to it must be licensed to all users, etc.
3-- Support reasonable compromizes like the one Amazon is backing. Not the best of all possible worlds, but as someone said, better than a poke in the eye with a broomhandle.
What are we waiting for?
Why patents are a good thing once the problem of obvious patents is solved (and how to solve that problem)
(1) Society benefits from a patent when the idea, design, machine, or program that is patented is (a) useful and (b) would not have been produced if it were not the patent system. Whether the thing patented is idea, design, machine, program is irrelevant if (a) and (b) hold.
(2) Society is harmed by a patent when the thing patented would have been produced anyway because the patent restricts its use.
(3) The problem with the recent explosion of patents is that many of them fall into the second category.
(4) The problem with scrapping the whole patent system or a class of patents, e.g. software patents, is that society would lose the benefit of patents of type. I expect that most slashdot readers would argue that this would be that this would be a good thing despite losing the benefit of type (1) patents. Perhaps this is true, but I think it is easy to jump to the wrong conclusion - when patents work, it is hard to see them working, when they don't work its is very obvious.
(5) The ideal situation would be to only have patents of type (1); the problem is finding an objective way to identify type (2) patents.
(6) Here is my solution. If you have something that you believe is worthy of a patent, you define the problem you have solved and send it together with a fee, say 10,000 dollars to the patent licensing body. The patent licensing body publicises the problem the patent applicant has defined for, say, 3 months. Anyone via the Internet is then invited, to come up with a solution to the problem. If someone comes up with the same solution as the patent applicant, the first person to do so gets to keep the 10,000 dollars and the solution enters the public domain. If no one comes up with the solution before 3 months have passed, then is not something obvious, so worthy of a patent. Clearly, people have a disincentive to apply for frivolous patents, e.g. one click, since there's a good chance they'd lose their money.
I hope this interests someone.
No, free software is used commercially, so it is not safe from software patents.
In any case, it is worth opposing software patents simply for their impact on *proprietary* software developers who are trying to develop new software and make money from it. The effect on free software is another item in a long list of Bad Things.
-- Ed Avis ed@membled.com