Those who are "religious" about the GPL might want to believe that the GPL is the epitome of virtue and purity, but its history indicates otherwise. It was designed, explicitly and intentionally, to be a "poison pill" for businesses.
The GPL has never been the foundation of a company that has been successful in the long run. Cygnus, for example, was only marginally profitable, and only after it adopted other licensing models (e.g. eCOS) and began to sell packaged software. Red Hat has lost many millions, and its own 10-Q statement (look it up on the SEC's Edgar system) warns that the company does not know how or whether its business model is viable. Eazel just laid of 60% of its staff (which is a shame; they're bright people who do good work!). If Ransom Love, CEO of Caldera, says that the GPL is open source's "weakest point," it only makes sense that he's determined that his business is also suffering due to the GPL.
Again, if you look at the history that led to the GPL (How many here have read the book Hackers? Or actually spoke with Richard Stallman at the time of development of the GPL as I did? At the time, he was much more frank about his intentions than he is today), you will recognize that it is not designed to make a viable business model possible. In fact, it's designed to undercut businesses, destroy their markets, and reduce programmers' wages (see the GNU Manifesto, which explicitly states this goal). I know that many folks who have bought into the idealistic (but misleading) language at the start of the GPL may find it hard to believe this -- so much so that I fear that this post will be moderated down in denial even though it's not a flame. But we really must wake up and smell the coffee: the GPL is anti-business and anti-programmer. There are other models that are better. The MIT X license and the BSD license are good examples. We should consider them.
I agree. Unfortunately, those "mafiosi" have teeth.... You might note that my remark two levels up from here was moderated down. Why? It wasn't a troll and expressed a valid opinion. The reason, of course, is that people have been blinded by religion -- to the extent that they'll moderate down comments which violate the "religious" tenets of the FSF.
--Brett Glass
OK, go back to your tape and play back what I really said, word-for-word, then type it in for us to read. Make sure it's word-for-word accurate. While you do
that, you'll find that "bend companies to my will" or "forcibly extract value from them" isn't in there.
It may take some time, but I'd be happy to dig it out. I've accurately paraphrased what you said. You also, as I recall, branded companies that used open source software to provide services as parasites, and said that you wanted to modify the GPL to extract valuable intellectual property from these companies as well.
Yes, Bruce, your animosity for commercial software companies. In your speech at the February 1999 LinuxWorld (which I'm fortunate enough to have on tape), you specifically stated that your goal was to bend corporations to your will -- to forcibly extract things of value from them. I believe that HP was very foolish to hire you, because you will, in fact, work against its interests given the chance.
Your "quid pro quo" is not a fair exchange; it is blackmail -- similar to an old fashioned "protection racket." Perhaps it would be a good idea to publish a transcript of your speech so that people can understand this.
One of the greatest scientists of our time, having done more for modern technology (and thus, btw, for the modern economy) that Microsoft will ever do, acknowledged the fact that he did so by being able to use the knowledge (what we now call "intellectual property") gathered by others.
In this passage, Linus confuses basic scientific knowledge with an invention, which is quite different. The discoveries of Newton, Einstein, Rutherford, Bohr, etc. were not inventions. They were discoveries of fundamental scientific principles. Such things are not and have never been patentable. By characterizing them incorrectly as "intellectual property," Linus starts his argument with an incorrect premise and thus renders it invalid. Don't get me wrong.... There are good arguments for limits on the scope and longevity of intellectual property,
but this one -- because it is built on a faulty premise -- is not valid.
If we assume that what Linus really meant is that we should be allowed to build modern technology on what came before, it would be an argument for the BSD license, or the public domain, and against the GPL (which prevents commercial programmers from standing on others' metaphorical "shoulders"). Microsoft, interestingly, seems to agree. It uses BSD-licensed code in its own products and does not condemn it in Craig's statement.
Corporations exist solely to maximize the benefit to their shareholders. The law says that they they are "fictional persons." But because they have a legally programmed-in, single-minded obsession with profit, they are -- in a very real sense -- golems. Noam Chomsky has a lot to say about this and the mixed bag of consequences that results from it.
In any event, corporations as they now exist do not exist to benefit consumers. If they did, who would invest in them? It would be desirable to teach them better manners, though.
I have to make a distinction between patents for physical processes outside of a machine that is already built which transform matter (as in the Diamond vs Diehr
case). These are fine by me since in order to violate the patent, you would have to physically take things outside of the computer and do stuff with them and put
materials into them. You could not violate that patent by buying a computer, taking it home, and typing some stuff into the computer. You would need to get raw
materials and other machines and process the raw materials in a certain way.
That is a very different situation than the typical software patent where I could buy a computer, take it home, type some things into it, and violate a patent by doing
that. That means the patent blocks expression.
Why the assumption that what one does with physical matter should be different than what one does with code?
Yes, it is true that one can use bits and bytes as a means of expression as well as for invention. But one can do either with physical matter, too! Give me a piece of wood and I might create a sculpture or design and build a novel machine. Give me a computer and I might write a moving story or invent a brilliant and useful new tool. The media are different. The processes are the same.
Invention in Cyberspace is no different from invention in the physical world, except that it is easier to avoid being bogged down by the logistics and expense of moving or shaping physical matter. The only limit is my brainpower, my gumption, and and ability to organize my work. When I code, I feel like a magician.... Steve Savitsky's wonderful poem/song The World Inside the Crystal, and Kathy Mar's brilliant performance of it, express the way I feel about this.
I do agree that because invention is so much easier in Cyberspace than in the physical world, the standards of novelty should be higher.... And, because technology advances so quickly, there are also good argument that the terms of patents should be shorter. But inventors in Cyberspace need protection -- both from companies that would squash them for profit and vindictive people that would sabotage them out of malice. It really looks as if Bruce wants to allow the FSF to do the latter.
Ah, the real meat. Couple an out-of-context quote of RMS with the claims of open-source software being "a weapon in a malicious agenda".
It's not an out-of-context quote. Stallman told me, in 1989, that he was trying to work out ways to destroy all commercial software companies, which he said were "evil," via GPLed code. One of his concerns, he said, was that patents could thwart his efforts to do this. He said that since he recognized the FSF approach to be somewhat "radical," and that programmers might recognize and fail to support its spiteful agenda, he would start a separate organization called the "League for Programming Freedom." This organization would focus only on trying to elminate software patents. In effect, it would therefore cause programmers who would not support the GPL or the FSF to further their ends indirectly, since not everyone who joined the LPF would realize that it was founded to support the FSF's radical aims.
Truly ruthless political scheming. But then, Stallman was (and is) driven. His sole desire is to wreak vengeance upon commercial software developers... and has been ever since his colleagues departed the AI Lab to earn a living as commercial programmers.
I'm not at all certain I agree with your assessment of Quarterdeck.
Microsoft introduced MEMMAKER into DOS 6.x solely as a reaction to the introduction of a similar utility into DR-DOS.
Your history is not correct. The product which Microsoft introduced to drive QEMM off the market was not MEMMAKER but rather EMM386. EMM386 was originally not as capable as QEMM, so Microsoft played dirty: it rigged Windows not to start if QEMM was installed. Instead, it gave an error message claiming that QEMM was not compatible with Windows. But, as columnist Jerry Pournelle reported in InfoWOrld, renaming QEMM to EMM386 (or any other name) foiled the detection code, allowed Windows to start -- and it ran just fine. (QEMM subsequently changed the name of its executable file to QEMM386 to sidestep this nastiness.) Most users, however, were spooked by the message and dropped QEMM for the free EMM386.
EMM386 was gradually improved to incorporate most of the features of QEMM and also those of DRI's memory manager.
To claim that Microsoft did this to get QEMM infers that DR-DOS had the same reason.
Not true. DR-DOS had a unique feature that not even Quarterdeck could implement without redesigning the OS: it was able to put parts of DOS into the High Memory Area (HMA). DRI wasn't out to hurt Quarterdeck and in fact promoted QEMM as a better memory manager than its own. I know whereof I speak here; I wrote a book on memory management and reviewed DR-DOS for InfoWorld (and later versions for PC World). I also verified the nasty Windows message while at InfoWorld.
It is true that Microsoft licensed MEMMAKER (which was not a memory manager itself but rather an optimizer that determined how to load programs to maximize the size of the DOS Transient Program Area) from Helix Software. (MEMMAKER was the optimizer from Netroom, but was initially one version behind the one that was sold with Netroom.) Netroom's optimizer was quite good; I discussed how it worked with the author (and the head of Helix), Mike Spilo. Mike told me, at the time, that he felt he had no option but to license the code, essentially for free, to Microsoft. It would squash his company and deny him access to vital technical information if he didn't. He'd been made an offer he could not refuse.
Then not that it mattered anyway, as at the time even with Microsoft's MEMMAKER one still had to purchase QEMM.
Not true. MEMMAKER worked with EMM386, which gradually incorporated all of the features of QEMM except for "Stealth" (a scheme in which EMS pages were used to implement a primitive form of transparent swapping). It would NOT work with QEMM. QEMM had its own optimizer.
A patent can't cover just a mathematical formula. According to Diamond v. Diehr, the precedent-setting case that paved the way for what are now called "software patents," they must cover an engineering solution that includes a mathematical formula.
I think of what's covered by software patents as "building a machine in Cyberspace." You're engineering a machine that works in a way that previous ones didn't, or does things that no other machine has done before. But instead of using physical parts, you're creating it in Cyberspace, using code. If the design is really novel and non-obvious, you should be rewarded just as if you'd machined metal or sawed wood to make it. And if you've made a serious contribution to technology, one of the things you should be protected against is the abuse of open source to prevent you from being rewarded for what you've done. Bruce takes an "open source uber alles" approach; he assumes that everything that's open source is automatically good. But open source is actually a powerful weapon for evil, as well. It can destroy businesses and lives if used in this way. Patents prevent that.
What Bruce fails to mention in his article is that patents defend against the abuse of open source as a weapon.
Let me explain why this is a concern. Suppose Company X makes Product A. Company Y competes with company X by making Product A and also makes
Product B, which provides the bulk of its income and finances the development of its competitive Product A. Company X can sabotage Company Y by fostering
the development of an open source equivalent of Product B so as to cut off the revenue Company Y needs to compete with it (and, perhaps, to survive). Its "air
supply" -- to use the word of Microsoft executive Jim Allchin -- has been cut off.
This isn't an abstract example. In the early days of the Windows environment, Microsoft destroyed Quarterdeck, which made DESQview, by giving away a free
knock-off of Quarterdeck's QEMM memory management software (which was used to support its GUI and multitasking development). The result: Quarterdeck,
without a "cash cow" equivalent to Microsoft's MS-DOS, could not compete. DESQview -- the best multitasking environment available for PCs at the time -- and
DESQview/X -- a brilliant GUI based on X Windows -- died because there was no money for their future development.
While Microsoft's free equivalent didn't happen to be open source, it very easily could have been. (In fact, had Microsoft not been rich enough to bankroll its
QEMM equivalent itself, it might have been the only practical way to go.) In any event, it always takes much less effort and money to develop a knock-off than the
original, because Quarterdeck blazed the trail by developing the technology and feature set, solving all of the knotty technical problems, and dealing with the
tradeoffs that are always inherent in new products.
Richard Stallman and others have specifically touted open source as a way of attacking companies they do not like (which, in the case of RMS, includes any
company that publishes commercial software -- even those which, unlike Microsoft, act ethically).
To prevent larger companies from copying their work and wiping them out -- and to prevent open source from being used in a predatory and unfair manner against them -- companies that develop new technology need to patent it. This is precisely
what patents are intended to do, and they're especially urgent in an age where open source can be abused to prevent people who honestly advance the state of the
art from being rewarded for their labors.
Richard Stallman once told me, personally, that one of his motivations for opposing software patents is that they would make it more difficult for the FSF to "wipe
out" (his words) commercial software. Stallman's statement underscores the danger.... And the need for software patents. We might argue about the optimal length
of these patents (there are many merits to the argument that 20 years is too long), and that some of them should not have been issued due to obviousness or prior art, but few of us with a sense of justice and fairness would say that open source
should be allowed to be used as a weapon in a malicious agenda.
If you've heard Bruce speak, or read his writing, you know that he shares some of RMS's animosity toward commercial software companies and frequently rattles
his saber, "demanding" that they forfeit their hard work. Could this be the reason he opposes software patents? Just food for thought.
What Bruce fails to mention in his article is that patents defend against the abuse of open source as a weapon.
Let me explain why this is a concern. Suppose Company X makes Product A. Company Y competes with company X by making Product A and also makes Product B, which provides the bulk of its income and finances the development of its competitive Product A. Company X can sabotage Company Y by fostering the development of an open source equivalent of Product B so as to cut off the revenue Company Y needs to compete with it (and, perhaps, to survive). Its "air supply" -- to use the word of Microsoft executive Jim Allchin -- has been cut off.
This isn't an abstract example. In the early days of the Windows environment, Microsoft destroyed Quarterdeck, which made DESQview, by giving away a free knock-off of Quarterdeck's QEMM memory management software (which was used to support its GUI and multitasking development). The result: Quarterdeck, without a "cash cow" equivalent to Microsoft's MS-DOS, could not compete. DESQview -- the best multitasking environment available for PCs at the time -- and DESQview/X -- a brilliant GUI based on X Windows -- died because there was no money for their future development.
While Microsoft's free equivalent didn't happen to be open source, it very easily could have been. (In fact, had Microsoft not been rich enough to bankroll its QEMM equivalent itself, it might have been the only practical way to go.) In any event, it always takes much less effort and money to develop a knock-off than the original, because Quarterdeck blazed the trail by developing the technology and feature set, solving all of the knotty technical problems, and dealing with the tradeoffs that are always inherent in new products.
Richard Stallman and others have specifically touted open source as a way of attacking companies they do not like (which, in the case of RMS, includes any company that publishes commercial software -- even those which, unlike Microsoft, act ethically).
To prevent open source from being used in a predatory and unfair manner against them, companies that develop new technology need to patent it. This is precisely what patents are intended to do, and they're especially urgent in an age where open source can be abused to prevent people who honestly advance the state of the art from being deprived of rewards for their labors.
Richard Stallman once told me, personally, that one of his motivations for opposing software patents is that they would make it more difficult for the FSF to "wipe out" (his words) commercial software. Stallman's statement underscores the danger.... And the need for software patents. We might argue about the optimal length of these patents (there are many merits to the argument that 20 years is too long), but few of us with a sense of justice and fairness would say that open source should be allowed to be used as a weapon in a malicious agenda.
If you've heard Bruce speak, or read his writing, you know that he shares some of RMS's animosity toward commercial software companies and frequently rattles his saber, "demanding" that they forfeit their hard work. Could this be the reason he opposes software patents? Just food for thought.
The Argentine law, as written, would not allow GPLed software. It reads:
From the date established by the Executive Power on, Public National Organizations mentioned in article 1 of this law, will not be allowed to
use programs that store data in non-public format, or with licenses which:
1.Imply any form of discrimination to people or groups.
2.Don't fulfill of the preceding Article 2.
3.Are specific or exclusive for only one product.
The GPL discriminates, explicitly and intenntionally, against commercial programmers. Richard Stallman says so in The GNU Manifesto, where he states that the intent of the GPL is to reduce programmers' wages and to destroy competition in the OS market (or any market where GPLed software takes root). Many proponents of the GPL actively (and proudly!) promote discrimination against commercial programmers as well. Whether or not one agrees with this sort of discrimination, it is -- any way you look at it -- discrimination. Against people and their products.
Therefore, the Argentine law, because it does not allow discrimination, would not allow the GPL's discrimination. Software published under the BSD or MIT X license would likely be OK, but the GPL would be right out.
Just read through the license, and it's unacceptable because it's viral. Using one sound clip from a work licensed under this license would require an artist who created something substantially new and valuable to give his or her work away for free, thus depriving him or her of compensation. It's the GPL all over again, but for musicians -- who can afford even less than programmers to be forced to forfeit the fruits of their labor. Just say no.
What Bruce (in his zeal) fails to understand, or hopes that companies won't recognize, is that the fact that the GPL has become so widespread makes it impossible, or at least extremely unwise, for inventors large and small to give up their patents. In fact, it gives them a greater incentive than ever before to sue for, and obtain, patents.
Here's why. Anyone who wants to make money from technology that involves software -- be it a large corporation or an individual developer -- is constantly threatened by the possibility that someone will come out with a GPLed equivalent of his product. It's not anywhere near as hard to come up with the GPLed copy as it is to come up with the original, which may have required millions or even billions of dollars in R&D! Once you can see how a company implemented a solution, it's trivial to copy its painstakingly developed techniques and carefully researched design decisions in a GPLed product, destroying the market and preventing the developer from even breaking even on development costs.
Patents provide a defense against the GPL by providing the developer with a temporary monopoly on his or her invention, and thus a guarantee that he or she will be able to recoup his or her investment.
Would any rational company -- including IBM or HP -- give up its one defense against having its business undermined and the value of its hard work destroyed? Not a chance. What's more, Bruce Perens' expositions of the topic (see his presentation at the February 2000 LinuxWorld Expo) contain thinly veiled threats and intimations of blackmail. This won't go over well with anyone -- especially companies such as IBM and HP. I expect them to say, "Sorry, Bruce, but we won't be bullied. And now that we see that your true goal is to undermine and hurt us, we'll make an extra effort to beef up our patent portfolios."
In short, I think that Bruce's efforts will not only fail but backfire.
Incidentally, your message above highlights the deception inherent in Stallman's writings. What he calls "Free Software" is not actually free at all. This "Doublespeak" is designed to deceive. GPLed software is not "free" in any sense of the word. In fact, it is extremely costly not only to programmers but to society because of the damage it does to innovation.
Do me a favor. IF nothing else, tell me why you feel you should be able to use everyone else's code for free?
I didn't say "everyone else's." What I did say was that the GPL is discriminatory -- intentionally so -- because it lets everyone use code in the way that most benefits him or her... except commercial programmers. The intended effect of this is to destroy the market for programmers' work, while at the same time preventing them from making a living by making incremental improvements on -- or integrating into their own work to avoid tedious reimplementation -- code that is already available to end users for free. Again, this is done with malice aforethought. For 20 years, RMS has had one goal in life: to destroy all commercial software companies out of spite for imaginary wrongs supposedly done to him many years ago.
If source code is made available for free to end users, it must also be available to programmers to improve and incorporate into their own work. Otherwise, it is not open source, because (again) it is not open to those who can make the best use of it. The purpose of the GPL is to turn open source into a weapon against commercial programmers and their livelihoods. To do this is just plain wrong... and programmers should not tolerate it.
The GPL removes incentives to innovation by destroying markets and making it impossible for programmers to be compensated for their efforts. All of us suffer because the state of the art is not advanced and better products are not available to us.
Very rarely will you see GPLed software that's innovative or novel; 99.99% of it merely copies commercial software. And once it guts a market, there's no reward to be had for advancing the state of the art.
This is what has happened in, for example, the area of compiler design. GCC has destroyed the market for C compilers, and compilers that are better, more efficient, or highly optimizing are now unavailable. Innovation in compilers is, alas, dead due to the GPL. Which is a shame, because all of us suffer due to the suboptimal code that GCC produces.
I'm a compiler writer; it was my first love and the field I always wanted to go into. But due to GCC and the GPL, I'd be foolhardy to attempt to publish a compiler or start a company that sold one. The GPL has killed that field of endeavor.
But since the GPL is the way it is, it coopts the software of lots of people who DON'T particularly believe in it but who find bits of GPLed software and don't really
care one way or the other.
Very insightful.... And very true. It appears that Linus Torvalds, when he started on Linux, fell into this category. He was looking for a license, and the GPL -- which was on the C compiler he was using -- was handy. He didn't know the ramifications of the GPL nor the agenda behind it.
Note, in the above message, the dogmatic use of the (capitalized) word "Free" -- as if it were a given that GPLed software were actually free. Because this is the most counterfactual element of the FSF's misleading dogma, GPL adherents seem to find it necessary to ram the falsehood home with capital letters.
The fact is that GPLed software isn't free at all. GPLed software is encumbered by a complex, multi-page license, rife with legalese. The reason why that license is so long -- instead of being short, sweet, and simple, like the BSD license -- is that it is designed to impose onerous restrictions on the use of the code. These restrictions are desgined to hurt programmers. They are intended to make it impossible for programmers to make a living by using, and building upon, the code in their own work. At the same time, the GPL encourages the distribution of the code, so as to undermine programmers' chances of ever making money from their hard work.
Ironically, the message above says,
Dont expect to have programmers react kindly to your stand point.
[sic]
The fact is that programmers, of all people, should be the ones most opposed to the GPL. The GPL is anti-programmer and anti-business, and (as I've mentioned in previous postings) is not a legitimate open source license. Its blatant discrimination against programmers is sufficient to disqualify it.
Apparently, you lack the guts to stand up against the unethical nature of the GPL, and therefore refuse to recognize that its terms are blatantly discriminatory. It's quite sad to see this.
In no way does the GPL "play fair." As Stallman himself says, the GPL is "not Mr. Nice Guy." It's a Faustian bargain designed to destroy programmers' livelihoods.
Stallman states, specifically, that hurting programmers' livelihoods is a purpose of the GPL. In The GNU Manifesto, Stallman writes:
What the facts show is that people will program for reasons other
than riches; but if given a chance to make a lot of money as well,
they will come to expect and demand it. Low-paying organizations
do poorly in competition with high-paying ones, but they do not
have to do badly if the high-paying ones are banned.
This is what Stallman advocates: banning commercial software and commercial
software companies. He furthermore wants to eliminate all programming
jobs which pay better than what is earned by a starving graduate student or
researcher. In short, he wants to destroy programmers' livelihoods.
Slashdot Mirror
The GPL has never been the foundation of a company that has been successful in the long run. Cygnus, for example, was only marginally profitable, and only after it adopted other licensing models (e.g. eCOS) and began to sell packaged software. Red Hat has lost many millions, and its own 10-Q statement (look it up on the SEC's Edgar system) warns that the company does not know how or whether its business model is viable. Eazel just laid of 60% of its staff (which is a shame; they're bright people who do good work!). If Ransom Love, CEO of Caldera, says that the GPL is open source's "weakest point," it only makes sense that he's determined that his business is also suffering due to the GPL.
Again, if you look at the history that led to the GPL (How many here have read the book Hackers? Or actually spoke with Richard Stallman at the time of development of the GPL as I did? At the time, he was much more frank about his intentions than he is today), you will recognize that it is not designed to make a viable business model possible. In fact, it's designed to undercut businesses, destroy their markets, and reduce programmers' wages (see the GNU Manifesto, which explicitly states this goal). I know that many folks who have bought into the idealistic (but misleading) language at the start of the GPL may find it hard to believe this -- so much so that I fear that this post will be moderated down in denial even though it's not a flame. But we really must wake up and smell the coffee: the GPL is anti-business and anti-programmer. There are other models that are better. The MIT X license and the BSD license are good examples. We should consider them.
--Brett Glass
I agree. Unfortunately, those "mafiosi" have teeth.... You might note that my remark two levels up from here was moderated down. Why? It wasn't a troll and expressed a valid opinion. The reason, of course, is that people have been blinded by religion -- to the extent that they'll moderate down comments which violate the "religious" tenets of the FSF. --Brett Glass
--Brett
Your "quid pro quo" is not a fair exchange; it is blackmail -- similar to an old fashioned "protection racket." Perhaps it would be a good idea to publish a transcript of your speech so that people can understand this.
--Brett Glass
If we assume that what Linus really meant is that we should be allowed to build modern technology on what came before, it would be an argument for the BSD license, or the public domain, and against the GPL (which prevents commercial programmers from standing on others' metaphorical "shoulders"). Microsoft, interestingly, seems to agree. It uses BSD-licensed code in its own products and does not condemn it in Craig's statement.
--Brett Glass
In any event, corporations as they now exist do not exist to benefit consumers. If they did, who would invest in them? It would be desirable to teach them better manners, though.
--Brett Glass
How about $39.95 retail?
--Brett
Yes, it is true that one can use bits and bytes as a means of expression as well as for invention. But one can do either with physical matter, too! Give me a piece of wood and I might create a sculpture or design and build a novel machine. Give me a computer and I might write a moving story or invent a brilliant and useful new tool. The media are different. The processes are the same.
Invention in Cyberspace is no different from invention in the physical world, except that it is easier to avoid being bogged down by the logistics and expense of moving or shaping physical matter. The only limit is my brainpower, my gumption, and and ability to organize my work. When I code, I feel like a magician.... Steve Savitsky's wonderful poem/song The World Inside the Crystal , and Kathy Mar's brilliant performance of it, express the way I feel about this.
I do agree that because invention is so much easier in Cyberspace than in the physical world, the standards of novelty should be higher.... And, because technology advances so quickly, there are also good argument that the terms of patents should be shorter. But inventors in Cyberspace need protection -- both from companies that would squash them for profit and vindictive people that would sabotage them out of malice. It really looks as if Bruce wants to allow the FSF to do the latter.
--Brett Glass
Truly ruthless political scheming. But then, Stallman was (and is) driven. His sole desire is to wreak vengeance upon commercial software developers... and has been ever since his colleagues departed the AI Lab to earn a living as commercial programmers.
--Brett Glass
EMM386 was gradually improved to incorporate most of the features of QEMM and also those of DRI's memory manager.
Not true. DR-DOS had a unique feature that not even Quarterdeck could implement without redesigning the OS: it was able to put parts of DOS into the High Memory Area (HMA). DRI wasn't out to hurt Quarterdeck and in fact promoted QEMM as a better memory manager than its own. I know whereof I speak here; I wrote a book on memory management and reviewed DR-DOS for InfoWorld (and later versions for PC World). I also verified the nasty Windows message while at InfoWorld.It is true that Microsoft licensed MEMMAKER (which was not a memory manager itself but rather an optimizer that determined how to load programs to maximize the size of the DOS Transient Program Area) from Helix Software. (MEMMAKER was the optimizer from Netroom, but was initially one version behind the one that was sold with Netroom.) Netroom's optimizer was quite good; I discussed how it worked with the author (and the head of Helix), Mike Spilo. Mike told me, at the time, that he felt he had no option but to license the code, essentially for free, to Microsoft. It would squash his company and deny him access to vital technical information if he didn't. He'd been made an offer he could not refuse.
Not true. MEMMAKER worked with EMM386, which gradually incorporated all of the features of QEMM except for "Stealth" (a scheme in which EMS pages were used to implement a primitive form of transparent swapping). It would NOT work with QEMM. QEMM had its own optimizer.
--Brett Glass
I think of what's covered by software patents as "building a machine in Cyberspace." You're engineering a machine that works in a way that previous ones didn't, or does things that no other machine has done before. But instead of using physical parts, you're creating it in Cyberspace, using code. If the design is really novel and non-obvious, you should be rewarded just as if you'd machined metal or sawed wood to make it. And if you've made a serious contribution to technology, one of the things you should be protected against is the abuse of open source to prevent you from being rewarded for what you've done. Bruce takes an "open source uber alles" approach; he assumes that everything that's open source is automatically good. But open source is actually a powerful weapon for evil, as well. It can destroy businesses and lives if used in this way. Patents prevent that.
--Brett Glass
Let me explain why this is a concern. Suppose Company X makes Product A. Company Y competes with company X by making Product A and also makes Product B, which provides the bulk of its income and finances the development of its competitive Product A. Company X can sabotage Company Y by fostering the development of an open source equivalent of Product B so as to cut off the revenue Company Y needs to compete with it (and, perhaps, to survive). Its "air supply" -- to use the word of Microsoft executive Jim Allchin -- has been cut off.
This isn't an abstract example. In the early days of the Windows environment, Microsoft destroyed Quarterdeck, which made DESQview, by giving away a free knock-off of Quarterdeck's QEMM memory management software (which was used to support its GUI and multitasking development). The result: Quarterdeck, without a "cash cow" equivalent to Microsoft's MS-DOS, could not compete. DESQview -- the best multitasking environment available for PCs at the time -- and DESQview/X -- a brilliant GUI based on X Windows -- died because there was no money for their future development.
While Microsoft's free equivalent didn't happen to be open source, it very easily could have been. (In fact, had Microsoft not been rich enough to bankroll its QEMM equivalent itself, it might have been the only practical way to go.) In any event, it always takes much less effort and money to develop a knock-off than the original, because Quarterdeck blazed the trail by developing the technology and feature set, solving all of the knotty technical problems, and dealing with the tradeoffs that are always inherent in new products.
Richard Stallman and others have specifically touted open source as a way of attacking companies they do not like (which, in the case of RMS, includes any company that publishes commercial software -- even those which, unlike Microsoft, act ethically).
To prevent larger companies from copying their work and wiping them out -- and to prevent open source from being used in a predatory and unfair manner against them -- companies that develop new technology need to patent it. This is precisely what patents are intended to do, and they're especially urgent in an age where open source can be abused to prevent people who honestly advance the state of the art from being rewarded for their labors.
Richard Stallman once told me, personally, that one of his motivations for opposing software patents is that they would make it more difficult for the FSF to "wipe out" (his words) commercial software. Stallman's statement underscores the danger.... And the need for software patents. We might argue about the optimal length of these patents (there are many merits to the argument that 20 years is too long), and that some of them should not have been issued due to obviousness or prior art, but few of us with a sense of justice and fairness would say that open source should be allowed to be used as a weapon in a malicious agenda.
If you've heard Bruce speak, or read his writing, you know that he shares some of RMS's animosity toward commercial software companies and frequently rattles his saber, "demanding" that they forfeit their hard work. Could this be the reason he opposes software patents? Just food for thought.
--Brett Glass
Let me explain why this is a concern. Suppose Company X makes Product A. Company Y competes with company X by making Product A and also makes Product B, which provides the bulk of its income and finances the development of its competitive Product A. Company X can sabotage Company Y by fostering the development of an open source equivalent of Product B so as to cut off the revenue Company Y needs to compete with it (and, perhaps, to survive). Its "air supply" -- to use the word of Microsoft executive Jim Allchin -- has been cut off.
This isn't an abstract example. In the early days of the Windows environment, Microsoft destroyed Quarterdeck, which made DESQview, by giving away a free knock-off of Quarterdeck's QEMM memory management software (which was used to support its GUI and multitasking development). The result: Quarterdeck, without a "cash cow" equivalent to Microsoft's MS-DOS, could not compete. DESQview -- the best multitasking environment available for PCs at the time -- and DESQview/X -- a brilliant GUI based on X Windows -- died because there was no money for their future development.
While Microsoft's free equivalent didn't happen to be open source, it very easily could have been. (In fact, had Microsoft not been rich enough to bankroll its QEMM equivalent itself, it might have been the only practical way to go.) In any event, it always takes much less effort and money to develop a knock-off than the original, because Quarterdeck blazed the trail by developing the technology and feature set, solving all of the knotty technical problems, and dealing with the tradeoffs that are always inherent in new products.
Richard Stallman and others have specifically touted open source as a way of attacking companies they do not like (which, in the case of RMS, includes any company that publishes commercial software -- even those which, unlike Microsoft, act ethically).
To prevent open source from being used in a predatory and unfair manner against them, companies that develop new technology need to patent it. This is precisely what patents are intended to do, and they're especially urgent in an age where open source can be abused to prevent people who honestly advance the state of the art from being deprived of rewards for their labors.
Richard Stallman once told me, personally, that one of his motivations for opposing software patents is that they would make it more difficult for the FSF to "wipe out" (his words) commercial software. Stallman's statement underscores the danger.... And the need for software patents. We might argue about the optimal length of these patents (there are many merits to the argument that 20 years is too long), but few of us with a sense of justice and fairness would say that open source should be allowed to be used as a weapon in a malicious agenda.
If you've heard Bruce speak, or read his writing, you know that he shares some of RMS's animosity toward commercial software companies and frequently rattles his saber, "demanding" that they forfeit their hard work. Could this be the reason he opposes software patents? Just food for thought.
--Brett Glass
From the date established by the Executive Power on, Public National Organizations mentioned in article 1 of this law, will not be allowed to use programs that store data in non-public format, or with licenses which:
1.Imply any form of discrimination to people or groups.
2.Don't fulfill of the preceding Article 2.
3.Are specific or exclusive for only one product.
The GPL discriminates, explicitly and intenntionally, against commercial programmers. Richard Stallman says so in The GNU Manifesto, where he states that the intent of the GPL is to reduce programmers' wages and to destroy competition in the OS market (or any market where GPLed software takes root). Many proponents of the GPL actively (and proudly!) promote discrimination against commercial programmers as well. Whether or not one agrees with this sort of discrimination, it is -- any way you look at it -- discrimination. Against people and their products.
Therefore, the Argentine law, because it does not allow discrimination, would not allow the GPL's discrimination. Software published under the BSD or MIT X license would likely be OK, but the GPL would be right out.
--Brett
--Brett Glass
Here's why. Anyone who wants to make money from technology that involves software -- be it a large corporation or an individual developer -- is constantly threatened by the possibility that someone will come out with a GPLed equivalent of his product. It's not anywhere near as hard to come up with the GPLed copy as it is to come up with the original, which may have required millions or even billions of dollars in R&D! Once you can see how a company implemented a solution, it's trivial to copy its painstakingly developed techniques and carefully researched design decisions in a GPLed product, destroying the market and preventing the developer from even breaking even on development costs.
Patents provide a defense against the GPL by providing the developer with a temporary monopoly on his or her invention, and thus a guarantee that he or she will be able to recoup his or her investment.
Would any rational company -- including IBM or HP -- give up its one defense against having its business undermined and the value of its hard work destroyed? Not a chance. What's more, Bruce Perens' expositions of the topic (see his presentation at the February 2000 LinuxWorld Expo) contain thinly veiled threats and intimations of blackmail. This won't go over well with anyone -- especially companies such as IBM and HP. I expect them to say, "Sorry, Bruce, but we won't be bullied. And now that we see that your true goal is to undermine and hurt us, we'll make an extra effort to beef up our patent portfolios."
In short, I think that Bruce's efforts will not only fail but backfire.
--Brett Glass
Incidentally, your message above highlights the deception inherent in Stallman's writings. What he calls "Free Software" is not actually free at all. This "Doublespeak" is designed to deceive. GPLed software is not "free" in any sense of the word. In fact, it is extremely costly not only to programmers but to society because of the damage it does to innovation.
I didn't say "everyone else's." What I did say was that the GPL is discriminatory -- intentionally so -- because it lets everyone use code in the way that most benefits him or her... except commercial programmers. The intended effect of this is to destroy the market for programmers' work, while at the same time preventing them from making a living by making incremental improvements on -- or integrating into their own work to avoid tedious reimplementation -- code that is already available to end users for free. Again, this is done with malice aforethought. For 20 years, RMS has had one goal in life: to destroy all commercial software companies out of spite for imaginary wrongs supposedly done to him many years ago.
If source code is made available for free to end users, it must also be available to programmers to improve and incorporate into their own work. Otherwise, it is not open source, because (again) it is not open to those who can make the best use of it. The purpose of the GPL is to turn open source into a weapon against commercial programmers and their livelihoods. To do this is just plain wrong... and programmers should not tolerate it.
Very rarely will you see GPLed software that's innovative or novel; 99.99% of it merely copies commercial software. And once it guts a market, there's no reward to be had for advancing the state of the art.
This is what has happened in, for example, the area of compiler design. GCC has destroyed the market for C compilers, and compilers that are better, more efficient, or highly optimizing are now unavailable. Innovation in compilers is, alas, dead due to the GPL. Which is a shame, because all of us suffer due to the suboptimal code that GCC produces.
I'm a compiler writer; it was my first love and the field I always wanted to go into. But due to GCC and the GPL, I'd be foolhardy to attempt to publish a compiler or start a company that sold one. The GPL has killed that field of endeavor.
--Brett
Very insightful.... And very true. It appears that Linus Torvalds, when he started on Linux, fell into this category. He was looking for a license, and the GPL -- which was on the C compiler he was using -- was handy. He didn't know the ramifications of the GPL nor the agenda behind it.
--Brett Glass
The fact is that GPLed software isn't free at all. GPLed software is encumbered by a complex, multi-page license, rife with legalese. The reason why that license is so long -- instead of being short, sweet, and simple, like the BSD license -- is that it is designed to impose onerous restrictions on the use of the code. These restrictions are desgined to hurt programmers. They are intended to make it impossible for programmers to make a living by using, and building upon, the code in their own work. At the same time, the GPL encourages the distribution of the code, so as to undermine programmers' chances of ever making money from their hard work.
Ironically, the message above says,
Dont expect to have programmers react kindly to your stand point. [sic]
The fact is that programmers, of all people, should be the ones most opposed to the GPL. The GPL is anti-programmer and anti-business, and (as I've mentioned in previous postings) is not a legitimate open source license. Its blatant discrimination against programmers is sufficient to disqualify it.
Apparently, you lack the guts to stand up against the unethical nature of the GPL, and therefore refuse to recognize that its terms are blatantly discriminatory. It's quite sad to see this.
Stallman did and does intend to destroy programmers' livelihoods. And says so -- not only in that document but repeatedly on many occasions.
In no way does the GPL "play fair." As Stallman himself says, the GPL is "not Mr. Nice Guy." It's a Faustian bargain designed to destroy programmers' livelihoods.
What the facts show is that people will program for reasons other than riches; but if given a chance to make a lot of money as well, they will come to expect and demand it. Low-paying organizations do poorly in competition with high-paying ones, but they do not have to do badly if the high-paying ones are banned.
This is what Stallman advocates: banning commercial software and commercial software companies. He furthermore wants to eliminate all programming jobs which pay better than what is earned by a starving graduate student or researcher. In short, he wants to destroy programmers' livelihoods.