news@helens.Stanford.EDU (news) (08/15/90)
gumby@Cygnus.COM writes: How ironic if it turns out that Free Software is the capitalist tool of the 90's! I'm sorry. "The Capitalist Tool" is a registered slogan of Forbes magazine. Free Software will have to find another slogan. Recant or prepare to be sued. brad@looking.on.ca (Brad Templeton) writes: Some would use a strictly utilitarian system, saying that patents exist only to encourage disclosure by bribing the inventor with a temporary monopoly. I think that is one reason for patents, but many also feel that there is something "right" about the inventor being rewarded. One utilitarian here. Patents dangle an economic carrot that serves two purposes. 1) Encouraging disclosure. 2) Fostering innovation. The intent is clearly stated in the U.S. Constitution: "The Congress shall have power ..... To promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries." Art 1.8. In the software arena today, circumstances are substantially different than when patent monopolies were introduced. 1) The pace of innovation is high. Even without patent protection, most of these "discoveries" would be made by someone else within a few years. 2) The lifespans of technologies are decreasing. 17 years of protection represents a much larger part of the entire useful life of a software technology than has historically been true of mechanical products. 3) University researchers discover many algorithms, and patents are not their main motivation. 4) Software can be a cottage industry. Unlike manufacturing, it requires comparably few resources to produce a product. Thousands of software patents are like land mines making it very dangerous ground for a small company. But given that monopolies are harmful to free competition, if software patents are not necessary for fostering innovation, why have them? brad@looking.on.ca (Brad Templeton) writes: We draw upon several ethics for this. The utilitarian is just one. But, as Yoda said, there is another. It is based on the concept that ownership should derive from creation -- that you own the results of your labours. Perhaps, but patents and UI copyrights not only protect against someone "taking" your idea. They also prevent someone else with the same original idea from using it. Even worse, they prevent someone with a new discovery, which builds upon the first, from using it, even if the new discovery is much more innovative and significant than the first. This stifles innovation. One could apply Brad's ownership arguments to scientific or mathematical discoveries, which often take years, or even decades, of work. Why shouldn't society allow such patents? Ethically, because granting the scientist ownership of the new idea would harm the rights of others more than it enhances the rights of the scientist, not to mention the "utilitarian" argument that it would destroy scientific progress. There's a continuum here. The question is where to draw the line. It's quite important, since laws and courts are very bad at drawing antialiased lines, i.e. handling the "gray" region near the boundary. Jim Helman Department of Applied Physics Durand 012 Stanford University FAX: (415) 725-3377 (jim@KAOS.stanford.edu) Voice: (415) 723-9127
brad@looking.on.ca (Brad Templeton) (08/15/90)
In article <JIM.90Aug14184448@baroque.Stanford.EDU> news@helens.Stanford.EDU (news) writes: >gumby@Cygnus.COM writes: > 2) The lifespans of technologies are decreasing. 17 years > of protection represents a much larger part of the entire > useful life of a software technology than has historically > been true of mechanical products. This is a tough one. While I have been a proponent of the reduced term patent for algorithms, it's hard to say what the term should be. For example, take public-key cryptosystems. Obvious idea? Hardly. Cryptology had been a major, war-funded area of research for many decades before public-key systems were created. They are very valuable -- an obvious win in many applications. They will let us do things in security we could not have done otherwise. Yet they were invented almost 17 years ago, and it is only now that the real use is coming. A 5 year patent would have left the authors with squat. Electronic spreadsheets? You think it's super obvious today, but I was one of the people demonstrating VisiCalc when it was launched in 1979 at NCC in New York. Yes, many people keyed in quickly, but it is amazing (in hindsight) to remember how difficult the idea was to get through to many experienced computer types. Not at all obvious at the time, it turns out. But in this case, a 5 year patent would have been plenty to make good money. >Perhaps, but patents and UI copyrights not only protect against >someone "taking" your idea. They also prevent someone else with the >same original idea from using it. Patents do this, that's true. And it's their biggest problem. I don't think the same applies to UI copyrights. At least not in the Lotus case. *Nobody* is coming up with the identical Lotus 1-2-3 menu tree independently. They are all sitting down with a copy of 1-2-3 and copying it. That is not innovation in the slightest. While I am not keen on it, I can see Lotus' case here pretty well. They designed a menu tree, which one can easily depict on a piece of paper. They made a successful claim that a software system that follows this exact tree is a derivative work based on that tree. Under the law -- which is pretty clear about derivative works -- I think they have a case. (This is not a "look and feel" case, however. It is a case based on the fact that a carefully designed menu tree is in of itself a work that can be protected.) Stopping the cloning of that menu tree harms nobody, except those who refuse to think of something on their own. The only reason to duplicate that tree is to take advantage of the immense effort Lotus has put into gaining and training customers who use that system. Not to innovate. I would not support anybody owning the concept of the menu tree in general, or any common menu system like 1-2-3's (with arrow keys, first-letters, etc.) But in this case it's not so much a style of interface that VP-Planner copied. It's copying the *exact* interface that I have no problem with forbidding. The only reason to do that is to do a knock-off. Yes, it's hard if you can't say, "here, use the same tree that you're used to." But that's what competing and innovating are all about. A similar approach applies to GUIs. I don't think anybody can claim that pointing at pictures is a new idea. But I do think people have a right to protect exact representations. Does this hurt standardization? I think not. Yes, it means you can't clone the popular interface. But look at the lessons of history. Standards come from poeople agreeing on them -- including the original inventor. The Compact Cassette is a famous example. Look at VHS vs. Beta. Look at the IBM PC vs. the PS/2. Time after time, the successful standards are the ones that the creators let out into the world either free or under reasonable terms. Those that people try to keep proprietary eventually die. You only get a standard that works when everybody is in agreement. So I say, let people protect their interfaces. It only dooms them, and promotes real innovation. -- Brad Templeton, ClariNet Communications Corp. -- Waterloo, Ontario 519/884-7473
dhinds@portia.Stanford.EDU (David Hinds) (08/15/90)
In article <1990Aug15.071911.3083@looking.on.ca> brad@looking.on.ca (Brad Templeton) writes: >In article <JIM.90Aug14184448@baroque.Stanford.EDU> news@helens.Stanford.EDU (news) writes: >>Perhaps, but patents and UI copyrights not only protect against >>someone "taking" your idea. They also prevent someone else with the >>same original idea from using it. > >Time after time, the successful standards are the ones that the creators >let out into the world either free or under reasonable terms. Those that >people try to keep proprietary eventually die. You only get a standard >that works when everybody is in agreement. > I think this is key - patents are not typically used as all-or-nothing weapons. Most patent disputes end in licensing agreements - not one company forbidding another to make a product period. I think IBM recently settled a patent dispute with Sun this way. Sun licenses its SPARC technology to gain an edge, by making it a standard. Using patents to guard a small niche and inhibit innovation generally isn't a sound economic strategy. Patents are supposed to encourage use of proprietary information, by guaranteeing that the patenter won't lose potential profits by making information public. I agree that 17 years is a long time for a software patent, though. Wouldn't it be fair to say, though, that because of the generally much much faster rate of innovation today compared to when patent laws were drafted, that ALL sorts of patents should be shortened? -David Hinds dhinds@popserver.stanford.edu
j@bucsf.bu.edu (James Allard) (08/16/90)
>>Electronic spreadsheets? You think it's super obvious today, but I was >>one of the people demonstrating VisiCalc when it was launched in 1979 >>at NCC in New York. Then you should remeber how "similar" 1-2-3's menu tree was to VisiCalc's. When I first sat down with 1-2-3, I got around pretty quickly with my knowledge of VC's functions, menus, etc.... I'd have to say that Lotus was "inspired".... >>*Nobody* is coming up with the identical Lotus 1-2-3 menu tree >>independently. ...... >>While I am not keen on it, I can see Lotus' case here pretty well. >>They designed a menu tree, which one can easily depict on a piece of >>paper. Yeah, but Lotus didn't really come up with their own either..... >>The only reason to duplicate that tree is to take advantage of the >>immense effort Lotus has put into gaining and training customers who use >>that system. Not to innovate. I agree wholeheartedly. It's jumping on the bandwagon. But aren't PC clone manufacturers just as (ir)responsible? I don't disagree with Brad's points. However, with all of this discussion about Lotus' lawsuit, I must admit that I'm surprised at how little VC was mentioned. Does anyone know anything about what VisiCorp is doing today, or where they stand in this debate? I'd be interested in hearing about it. One more (small) point: Since there is a finite number of say, two-stoke keyboard sequences on a 101 key keyboard, say n, then is it impossible for n+1 manufacturers to produce applications w/o fear of lawsuits? I really find this whole issue tough to swallow. Just a curious kid, I don't claim to know anything (much). -- =-+-= =-+-= =-+-= =-+-= =-+-= =-+-= =-+-= =-+-= =-+-= =-+-= =-+-= =-+-= =-+-= J. Allard A funny thing about regret is 109 Peterborough St. #12A That it's better to regret something Boston, MA 02215 You *have* done (617) 424-8818 Then to regret something You *haven't* done And by the way..... =-+-= =-+-= =-+-= =-+-= =-+-= =-+-= =-+-= =-+-= =-+-= =-+-= =-+-= =-+-= =-+-=
jim@baroque.Stanford.EDU (James Helman) (08/16/90)
brad@looking.on.ca (Brad Templeton) writes:
Stopping the cloning of that menu tree harms nobody, except those who
refuse to think of something on their own.
This is another case where the benefits of UI copyrights to the
originator of a design must be balanced against the harm they cause to
users and competitors.
Is it fair to Lotus that some little upstart could come along and
knock-off a copy of Lotus and sell it for less? Probably not. But in
practice, it takes more than a cheaper knock-off to compete against a
established product with a good reputation. A competing package would
have to offer some significant benefit over Lotus before people would
switch, even to a look-alike product. But if the competing product is
incompatible, people will be very reluctant to switch, even if it is
significantly better. If look-alike products are prohibited, Lotus
has less incentive to stay ahead of the pack in innovation.
The only reason to duplicate that tree is to take advantage of the
immense effort Lotus has put into gaining and training customers who
use that system. Not to innovate.
Actually, customers have invested much more time and effort in
learning an interface than the vendor has in training them. This is
my biggest problem with UI copyrights. I HATE LEARNING NEW
INTERFACES. Or more precisely, I hate switching interfaces. For a
while, I was going between SunView and X10/uwm on Suns, NeWS on SGIs,
old DECWindows on VMS Vaxen. After getting quick and mouse-agile on
one, I'd go to another and make lots of mistakes. I felt like a rat
in an experiment in which the experimenter kept switching the food
reward and the shock punishment. After enough of this dissonance, I
slowed down a lot and became averse to mice until I got X/twm running
on everything.
None of these systems offered any huge UI advantage over another; they
were just different. I don't think UI difference for its own sake
should be encouranged. It promotes stress and causes user errors.
For example, according to one report, one of last years tanker
accidents happened because the operator was accustomed to a control
that worked the opposite as the one on the ship.
Jim Helman
Department of Applied Physics Durand 012
Stanford University FAX: (415) 725-3377
(jim@KAOS.stanford.edu) Voice: (415) 723-9127
gumby@Cygnus.COM (David Vinayak Wallace) (08/16/90)
Date: 15 Aug 90 16:48:27 GMT From: dhinds@portia.Stanford.EDU (David Hinds) Most patent disputes end in licensing agreements - not one company forbidding another to make a product period. But this means I can't give my software away, since I will have to keep track of who gets copies in order to pay a licensing fee! At least I can still circulate the source if I want; patents themselves are in the public domain. If companies want to keep stuff secret, fine. Let them go back to using trade secrets (not these stupid "trade secrets" now where companies distribute .h files and then claim that they are valuable trade secrets and that you mustn't read them). Trade secrets allow me the opportunity to re-invent with impunity, and they don't amount to a government subsidy to the software industry.
brad@looking.on.ca (Brad Templeton) (08/16/90)
Actually, one thing that I think will be very interesting in the future is the concept of the free clone. Nobody has really done it yet, but some day it will happen, either from the GNU project or elsewhere. A drop-in, fully compatible software clone that is free. Now it won't wipe out the maker of the original, but neither will it go away. The middle ground is hard to predict. Actually, many interesting factors come into play. For example, what about deliberate corproate sabotage? What if Microsoft (hypothetical example, I do not imply MS would do this) were to make a *really good* 1-2-3 clone, and then release it as freeware. And push it -- include it free with every DOS or Windows etc. Free with Excel, too. That could do *serious*, perhaps crippling damage to Lotus, establishing MS as the big contender in software. It would hurt Excel, but not the way that it would hurt 1-2-3. Would this be an anti-trust violation, perhaps? The anti-trust laws stop people from deliberately giving away stuff to kill the competition -- might that apply to software? I actually think a company should get some protection from this. And I do understand what has been said about frustration in changing interfaces. The answer is to get people to standardize -- willingly. They *will* do it if they are smart. -- Brad Templeton, ClariNet Communications Corp. -- Waterloo, Ontario 519/884-7473
prisoner@pawl.rpi.edu (Allen S. Firstenberg) (08/16/90)
j@bucsf.bu.edu (James Allard) writes: >I don't disagree with Brad's points. However, with all of this discussion >about Lotus' lawsuit, I must admit that I'm surprised at how little VC was >mentioned. Does anyone know anything about what VisiCorp is doing today, >or where they stand in this debate? I'd be interested in hearing about it. VisiCorp? Yeah. They got into financial trouble some years back. They were finally purchased by some company in Massachusets... um... what was it's name? Lotus or something like that. -- prisoner@pawl.rpi.edu "Do you know what this means?" prisoner@xrdlab1.mat.rpi.edu "Yes, It means I'm free." prisoner@rpitsmts.bitnet "No" -Les Miserables
brad@looking.on.ca (Brad Templeton) (08/16/90)
In article <S7%%AP=@rpi.edu> prisoner@pawl.rpi.edu (Allen S. Firstenberg) writes: >j@bucsf.bu.edu (James Allard) writes: > > >>I don't disagree with Brad's points. However, with all of this discussion >>about Lotus' lawsuit, I must admit that I'm surprised at how little VC was >>mentioned. Does anyone know anything about what VisiCorp is doing today, >>or where they stand in this debate? I'd be interested in hearing about it. > >VisiCorp? Yeah. They got into financial trouble some years back. They >were finally purchased by some company in Massachusets... um... what >was it's name? Lotus or something like that. > You did forget its name. Lotus never purchased VisiCorp. VisiCorp still exists as a shell today, to the best of my knowledge, though it does very little if anything. Lotus purchased Software Arts, which was the company that developed VisiCorp. Bob Frankston went to work for Lotus, Dan Bricklin formed his own company to sell tools called Software Garden. It's a long and sordid story. Look it up in a history of the software industry some day. -- Brad Templeton, ClariNet Communications Corp. -- Waterloo, Ontario 519/884-7473
abbott@aerospace.aero.org (Russell J. Abbott) (08/18/90)
In article <GUMBY.90Aug15120821@Cygnus.COM> gumby@Cygnus.COM (David Vinayak Wallace) writes: > ... >At least I can still circulate the source if I want; patents themselves >are in the public domain. Is this true? What are the implications of this? Can one write, distribute, perhaps even sell a program that uses a patented algorithm without violating the patent as long as one distributes the program in source form? Can one argue that when a patented algorithm is distributed in source form one is in some sense only discussing a publicly published patent--that it is not until the code is compiled, loaded, and executed that the patent is violated? If one wanted to use a patented algorithm as a component in one's software product, could one include some source code that uses the algorithm as well as some code that uses a different (perhaps slower) algorithm and leave it up to the customer to work out a licensing agreement with the patent holder if he wishes to use the patented algorithm? I really like the implications of this! If one of the points of a patent is to make the patented material public knowledge, then presumably there aren't any limitations on distribution of the patent documentation--including the source code! In effect this would mean that patenting an algorithm would virtually make it shareware. I love it! -- Russ Abbott@itro3.aero.org