The problem is that instead of screwdrivers, they're teaching the kids to program small robots that drive over and drive in the screw, and requiring that all teaching use these instead of screwdrivers.
I'm not defending the authors of the article, but...
1. this article wasn't written by rdbms people, but rather by column database people. There's nothing traditional or relational about their background. A column database is a particular way of implementing relational databases. David DeWitt (faculty homepage, wikipedia entry) is best known for object-relational work, which is in the traditional relational area. Michael Stonebraker, the other author, is probably the best known relational database person living today (though C.J. Date might be another candidate).
HIV can be transmitted from mother to child, and through poor hygiene (getting other's blood in your own small cuts), and through the (medical or drug) use of needles that aren't properly sterilized. HIV infection is not any clear indicator of sexual activity.
Not that I would be surprised at sexual activity at 15 - as far as I remember, the average sexual debut age here (Norway) is 15 for girls (17 for boys). It is at least clear that sexual debut before 15 is quite common.
Let's not kid ourselves - using the word "stealing" for "copyright infringement" means we can't think clearly about the case, as we get a cognitive mess up between two different things.
It is also clear that attitudes follow behaviors - those that do copyright infringement will tend to think that is more OK than they did before they did infringement, those that release closed software/films will think that "protection" of their copyrighted material ("property" is another cognitive distorter) is important and "their right", etc.
The issue with copyright laws is what effect do they have on people's behavior, in total. This is a combination of benefits - people creating more, some people feeling that they get rightfully compensated, etc. They also have some negative effects - for instance, the feeling that "copyrighted material" is property, when it has been released as part of culture and is actually part of other people's minds. Which the originator isn't paying rent for. Or another negative effect: Those that break copyright law end up disrespecting laws overall. Or another: They block people from creating derivative works. Or another: They end up closing in a lot of stuff that would otherwise be public domain, and where the author has no monetary interest - it's just inconvenient. Or another: Their enforcement end up with draconian policies hitting everywhere.
Personally, I am in favor of copyright, assuming the right limitations. I think that it is reasonable to be able to block commercial use for a limited time, against proper release of the materials afterwards. A reasonable limited time for software is probably in the range of 2 to 5 years, with "proper release of materials afterwards" meaning source code/version control dump release, with build files etc. For entertainment, reasonable a copyright term is probably about 5 years.
The net result of freeing private copying is that money would have to be made elsewhere: Movies would have to make their money at the box office and from the benefits of buying DVDs *beyond access to the actual movie data*, music would have to give benefit *beyond the actual audio data*, etc. Easy delivery of the data is one such benefit; at the right price, that's value. Packaging you can put on the shelf is another.
This might kill really expensive movie production. I feel this is OK - there are a ton of good movies made cheaply, movies that mostly languish due to little marketing. I believe people switching to these would result in people that are as happy as they are today - and possibly people that are more reflected - and as a such would be at least net neutral to society, and if this result in less of society's resources going into movie production with the same (or higher) benefit coming out: Net positive.
I hadn't thought of the last one until now - interesting.
There are three kinds of people who fund open source development:
Those selling complementary products / services.
Those who actually need the software.
Those developing in their spare time to pad their CVs.
Add those that do it for a hobby / because they are part of a community that they want to give back to / participate in. That's the largest part as far as I can tell. Or maybe I've just been lucky enough to be part of some very good communities?
That doesn't count because the syntax is a special case just for those functions.:) No, it isn't. You can do this for any function by using perl prototypes.
Of course, as always, perl has no reasonable parameter passing. And it doesn't work with method calls, as prototypes are disregarded for method calls. You can reasonably discount it for that - but in theory, it isn't special for map/sort.
But... yes... back to the Ruby hacking... Good choice:)
apt-get or package managers in general don't help much for proprietary software, which needs to work on as many versions as possible. And it only work for free software because somebody is putting in the maintenance time.
I don't program in Python, yet I don't see the tab differences as being a common problem for system administrators: In my experience, few system administrators are dumb enough to adjust tab size from the default 8. Programmers often do, because they often work so much in a text editor centered at a single place (their primary development machine) that funky adjustments are reasonable; system administrators don't, as they usually work on many different machines.
I've read these several times, though it's a few years since the last time. I just reread "Why software should not have owners" and "Why software should be free" since these were the ones that seemed relevant to the particular arguments. I didn't find them so, as I find the issues I detail below to counter much of the arguments put forth. The argument that DOES find some sympathy with me is the psychosocial one - I just have a slightly different bent on it. My thought is that the optimal frame is "Most generic software is free, and when you need something that is so specialized for you that nobody has had the time to make a free version yet, you pay for the specialization". In other words, the view of purchasing software as a contribution to development. This relies on less draconian enforcement and large amounts of really free (PD/BSD) software, though.
My issues with the essays:
For "Why software should not for have owners", in this the argument against the creation of value for end users is based on the notion that proprietary software is bad therefore having incentives that create more proprietary software is bad. The "is bad" part is based on incidental issues (too draconian enforcement, delivery as binaries, etc).
In "Why software should be free", Stallman assumes the edge case (proprietary software vs no software) does not exist. He calls the argument about the edge case - which is what is interesting - "begging the question". At least today, and for me since the late 80s or early 90s, it is obvious that there can be free software for the mainstream things, where "mainstream" is defined by programmer interest. He goes on to create unreal ethical constraints (a la "if giving your salary above $20,000 per year to the Red Cross is better for society overall than keeping it, all ethical programmers would give it to the Red Cross"). He constructs a bunch of arguments for how keeping software locked up is bad for society, with these three as a core:
Fewer people use the program.
None of the users can adapt or fix the program.
Other developers cannot learn from the program, or base new work on it.
With the exception of "learn from the program" and a replacement of "None" with "Fewer, this applies to locking code down with the GPL as compared to BSD-style licenses (or as I link to think of them, public domain style licenses).
I've also heard Stallman speak on copyright. Here, I noticed what I see as a strong inconsistency: He wants copyright to be optimal for everything but functional works, and then he goes off with a completely different argument for functional works: "Can be made to work".
If you have something specific that you think counters my views, by all means point it out. Just know that they're worked out over more than 15 years, including having been a fan of many sides of GNU and later having changed my views based on careful thinking about what influence what. I've been both much more pro-GNU than I am, followed by an anti-GNU period, followed by my presently more relaxed stance towards them. I still dislike their rhetoric and use of force; and I feel that if we were to use force, then a more appropriate use of force would be to force release of source code after a period of time (e.g, 3 years) instead of immediately.
Open source advocates disagree, seeing software development not as a social activity with ethical ramifications
That we have different priotities than you doesn't mean that these are non-ethical. I have priorities in delivering value to end users. Software and licenses is just a tool. I see the GPL as non-ethical license because it denies end users freedom to choose by denying the building of some of the things the end users would want to choose between.
In practice, this denies marginal end user groups the freedom to get rid of risk by purchasing pre-customized software off-the-shelf, where the risk of doing the customization has been taken by a software developer (including time risk and acquisition cost risk - ie, having the entire group go together and split the costs of development is not a viable replacement).
In other words, I see proprietary derivates of free software as creating a social benefit. I see it as a social activity.
Anyway, What is scarce in a free software economy is developer time. To increase freedom through free software, we need to increase choice - which means to produce more software, with free software having a higher value than proprietary software. However, proprietary software has a higher value than no software at all, and proprietary software derived from free software has a large chance of feeding back to the free software it is derived for, thus adding value at two points. This means that the freedom to create proprietary derivates creates values in two forms: Both the software (and thus utility to users) created purely proprietary, and the developer time it gives back to the free software original. In my experience, the deliveries back are substantial - for instance, both the SCSI subsystem and the netgraph subsystem of BSD was given back by proprietary derivates. When I did a proprietary derivate of FreeBSD (delivering an ISDN router/web proxy/UUCP mailer to customers), about 90% of the changes we did were contributed back. The last 10% - and the possibility of keeping other changes if we had needed to - worked as pay for the rest, and would only have been relevant for a direct competitor (as they munged FreeBSD as a general system - I messed up the tty code to make other things work).
The GPL, by contrast, gets extra developer time the following ways (and that's all I have found - I would be glad to hear of more):
Forcing developers to use an inferior product and improve on it or single end users to use an inferior product and pay for development of it. (When the commercial alternative is blocked and the programmer would have preferred to be an end user or the end user would have preferred to buy off-the-shelf.)
Manipulating developers into thinking the GPL is about extra freedom.
Creating and then handling developers fear of being "exploited" by somebody else taking your code, adding value, and selling the extra value ("somebody is earning money from MY CODE").
Misunderstandings of how the GPL applies / community forcing release when the developer did not want it and would not have done the development if (s)he understood how the license worked (examples are MOSIX and the original Objective C frontend from NeXT.)
Used as a product "demo" license, where the original developer allows companies to buy themselves out of the GPL, and the money from these buyouts fund development.
Potentially some sort of "equanimity" thinking on the part of developers allowing them to feel better about giving out code. I have heard this as a rationale for using the GPL only by Bruce Perens, and only when I challenged him on GPL use. I personally feel this is a rationalization from having already used the license a lot and feeling a need to have some reason, still, I'll leave it in for completeness.
Forcing code release when somebody thought they could get away with appropriating GPL software as a basis for their commercial product and it turne
There's a bug in your dream routine. You should dream of the attacker taking your gun or of your kids finding it and having an accident, or of you having an accident, or of somebody you are close to getting the gun and killing somebody else. Because this is what MOSTLY happen with such guns. More than you shooting an attacker.
Eivind, who has nothing against guns per se, but think the way people see guns in the US (including "self defence" with a gun as an idea) is the most likely source of all the horrible gun violence in the US.
As far as I've understood this, the US is forcing those things into the treaties and then saying "We have to do them because of the treaties". Nobody but the US copyright industry seems to want them, and underhanded means (push it through the diplomats quietly) are used to get them into WIPO etc.
If the purpose of modern copyright is to encourage the creation of artistic works, I can guaran-fucking-tee you that taking away people's own creative work is pretty much the worst possible way to achieve that. Since granting the copy privileges make it impossible for other people to create the creative works they want (based on the first work), it's clear that giving the privileges is discouraging in some situations.
What is happening isn't taking away the work, anyway. We're saying that the originator doesn't have infinite privileges *when (s)he chooses to make the work public*. The moment somebody make a work public, it is only partially their work - because part of the value of the work is the fact that the work is in the minds of others (AKA a part of culture). Thus part of the value of the work belongs to those people. If too large privileges are granted to the originator of the work, then the trade goes bad.
Evolutionary reason for dreaming, it seems like a silly thing to evolve a period of a beings life where they body goes into paralysis just so they don't kill themselves from acting lucid imagery, the fact the dreams gave us a survival advantage would explain the tradeoff of the paralysis during the night.
This assumes that all elements of life in this reality resolve down to questions of evolutionary theory, which I think is false. --I tend to think that we are not living in a closed system; that there are a LOT of outside forces at work which dramatically affect the human species and which have little to do with natural selection, --that and the rules which govern our reality are infinitely more complex than is currently understood. I find this very overbearing. You're just listing vague claims and attacking people with them, using Ericssonian language tricks to bring about a hypnotic trance and a vague sense of agreement because you allow people to fill in meaning for themselves.
Dirty arguing. Make some concrete, testable claims. SAY something, something REAL. Not just something that use language to manipulate. Manipulative language is a dirty game, and some of us happen to know enough to spot it.
If you're not playing this game on purpose, you've probably hung around too many "alternative" groups and got it by assimilation - many people there have it naturally, and others increase the force of it through NLP training.
What a shame that the 'experts' had to torture innocent rats to come up with their hypothesis. I would have thought they could have worked out whether their hypothesis was possible by common sense.
Rats' dreams are no doubt similar to our own, which is why torturing them and using them as 'things' is wrong... Don't worry, due to the innocence issue they've switched to using lawyers.
In practice, we do not add quite that much, because much of that (I seem to remember about 2/3s) is disappearing somewhere we do not understand. I personally am quite scared by that, too - we're dependent on some carbon sink we do not understand how works, so we have no idea when or if it will become full...
We can deal with the production of power in the day and consumption at night by using power storage. This can presently be done at about 80% efficiency, through the use of water storage (you pump water up into a reservoir when you have surplus power, and release it when you need to draw power).
The difference in consumer voltage between Europe, Japan and the US is a non-issue - we transport electricity at a much higher voltage, and then transform it down close to the point of use. The same isn't quite true for frequency - it is synced at 50/60Hz in the grid - but there are production facilities in operation that produce it at a different frequency and convert it to the grid frequency using a frequency changer. You can read more about in Wikipedia's utility frequency article.
The main problem with interconnecting the continents is the power loss associated with long distance transmission. As far as I understand, this makes interconnection impractical at the moment - local storage (as in the reservoirs described above) being more economical. Superconductors may some day change this.
They already do. And it may be a good thing; the value of the norms is that they make all of us able to function, and removing weight from them allow people to disrupt overall functioning of everybody else.
Of course, like everything else, it is a tradeoff. I have a fair amount of tolerance for people's behavior, and basically feel that all behavior is OK as long as it doesn't hurt people. However, I have fairly little tolerance for people hurting others because "they are brilliant and should be allowed to act that way". It is possible to both be brilliant and learn to treat others well - and I feel that the latter is reasonable to ask.
The problem here is that you don't understand the purpose of on-line letter journals, the purpose of peer review, or the fallibility of scientific publications. You are so scientifically naive that you actually think it's worth remarking that a peer-reviewed paper might be wrong. And you're so arrogant that you think that a bunch of Slashdot geeks with no experience in sleep research or biology can spot methodological problems better than the peer reviewers. There seems to be a little mistake in your logic here? You seem to assume that spotting extra methodological problems in one particular paper is the same as generally spotting methodological problems. The reviewers may be good at spotting methodological problems in general and still miss one; and us other people that are generally less qualified may be able to spot particular methodological errors that have passed the reviewers. For instance, I spotted a methodological error in a death rate research paper (and contacted the authors about this and got feedback from them acknowledging the issue) and I am generally not at all qualified in the area. I just happened to be able to notice this error because I had competence from another area that happened to make me look at other aspects of their data.
On a side note, I tend to be skeptical of all negative-to-radiation research with Swedish researchers, because my impression is that they have a history of debunked claims, sometimes with gross method errors, sometimes with minor ones, and usually with a bunch of publicity when they come with the claims, and little publicity when the claims are debunked.
As for the work you describe, it bears about as much relation to real-world operating systems as anti-gravity research does to wheels. Yeah, sounds wonderful, it'd be nice if it worked, but there are some fundamental reasons why it won't. And there are some fundamental arguments I have for why you're wrong. If you share your fundamental reasons maybe I'll share my arguments against them;) (Or maybe I'll applaud and say whoa, I hadn't thought of that...)
Example of the problem: Jordan K. Hubbard switched form the GPL to the BSD license after having had to rewrite code he had mostly written himself but had taken in contributions to. From what I understood from him, 90+% his - but he hadn't tracked the contributors/contributions and couldn't precisely identify which was his code and which was contributed, so he had to rewrite all of it.
I personally do not use the GPL, and have had the advantage of being able to reuse the codebase I know (FreeBSD) for a kernel-tuned ISDN-router/Squid proxy/UUCP mail system, and thereby be able to spend more time on it and contribute more code back to it. With the GPL, the projects I did most likely just wouldn't have gotten done, so no derivate - and less contribution.
Slashdot Mirror
Eivind.
Eivind.
1. this article wasn't written by rdbms people, but rather by column database people. There's nothing traditional or relational about their background. A column database is a particular way of implementing relational databases. David DeWitt (faculty homepage, wikipedia entry) is best known for object-relational work, which is in the traditional relational area. Michael Stonebraker, the other author, is probably the best known relational database person living today (though C.J. Date might be another candidate).
Eivind.
Not that I would be surprised at sexual activity at 15 - as far as I remember, the average sexual debut age here (Norway) is 15 for girls (17 for boys). It is at least clear that sexual debut before 15 is quite common.
Eivind.
Let's not kid ourselves - using the word "stealing" for "copyright infringement" means we can't think clearly about the case, as we get a cognitive mess up between two different things.
It is also clear that attitudes follow behaviors - those that do copyright infringement will tend to think that is more OK than they did before they did infringement, those that release closed software/films will think that "protection" of their copyrighted material ("property" is another cognitive distorter) is important and "their right", etc.
The issue with copyright laws is what effect do they have on people's behavior, in total. This is a combination of benefits - people creating more, some people feeling that they get rightfully compensated, etc. They also have some negative effects - for instance, the feeling that "copyrighted material" is property, when it has been released as part of culture and is actually part of other people's minds. Which the originator isn't paying rent for. Or another negative effect: Those that break copyright law end up disrespecting laws overall. Or another: They block people from creating derivative works. Or another: They end up closing in a lot of stuff that would otherwise be public domain, and where the author has no monetary interest - it's just inconvenient. Or another: Their enforcement end up with draconian policies hitting everywhere.
Personally, I am in favor of copyright, assuming the right limitations. I think that it is reasonable to be able to block commercial use for a limited time, against proper release of the materials afterwards. A reasonable limited time for software is probably in the range of 2 to 5 years, with "proper release of materials afterwards" meaning source code/version control dump release, with build files etc. For entertainment, reasonable a copyright term is probably about 5 years.
The net result of freeing private copying is that money would have to be made elsewhere: Movies would have to make their money at the box office and from the benefits of buying DVDs *beyond access to the actual movie data*, music would have to give benefit *beyond the actual audio data*, etc. Easy delivery of the data is one such benefit; at the right price, that's value. Packaging you can put on the shelf is another.
This might kill really expensive movie production. I feel this is OK - there are a ton of good movies made cheaply, movies that mostly languish due to little marketing. I believe people switching to these would result in people that are as happy as they are today - and possibly people that are more reflected - and as a such would be at least net neutral to society, and if this result in less of society's resources going into movie production with the same (or higher) benefit coming out: Net positive.
I hadn't thought of the last one until now - interesting.
Eivind.
- Those selling complementary products / services.
- Those who actually need the software.
- Those developing in their spare time to pad their CVs.
Add those that do it for a hobby / because they are part of a community that they want to give back to / participate in. That's the largest part as far as I can tell. Or maybe I've just been lucky enough to be part of some very good communities?Eivind.
Of course, as always, perl has no reasonable parameter passing. And it doesn't work with method calls, as prototypes are disregarded for method calls. You can reasonably discount it for that - but in theory, it isn't special for map/sort.
ButEivind.
Eivind.
Eivind.
Eivind.
My issues with the essays:
For "Why software should not for have owners", in this the argument against the creation of value for end users is based on the notion that proprietary software is bad therefore having incentives that create more proprietary software is bad. The "is bad" part is based on incidental issues (too draconian enforcement, delivery as binaries, etc).
In "Why software should be free", Stallman assumes the edge case (proprietary software vs no software) does not exist. He calls the argument about the edge case - which is what is interesting - "begging the question". At least today, and for me since the late 80s or early 90s, it is obvious that there can be free software for the mainstream things, where "mainstream" is defined by programmer interest. He goes on to create unreal ethical constraints (a la "if giving your salary above $20,000 per year to the Red Cross is better for society overall than keeping it, all ethical programmers would give it to the Red Cross"). He constructs a bunch of arguments for how keeping software locked up is bad for society, with these three as a core:
I've also heard Stallman speak on copyright. Here, I noticed what I see as a strong inconsistency: He wants copyright to be optimal for everything but functional works, and then he goes off with a completely different argument for functional works: "Can be made to work".
If you have something specific that you think counters my views, by all means point it out. Just know that they're worked out over more than 15 years, including having been a fan of many sides of GNU and later having changed my views based on careful thinking about what influence what. I've been both much more pro-GNU than I am, followed by an anti-GNU period, followed by my presently more relaxed stance towards them. I still dislike their rhetoric and use of force; and I feel that if we were to use force, then a more appropriate use of force would be to force release of source code after a period of time (e.g, 3 years) instead of immediately.
Eivind.
Open source advocates disagree, seeing software development not as a social activity with ethical ramifications
That we have different priotities than you doesn't mean that these are non-ethical. I have priorities in delivering value to end users. Software and licenses is just a tool. I see the GPL as non-ethical license because it denies end users freedom to choose by denying the building of some of the things the end users would want to choose between.
In practice, this denies marginal end user groups the freedom to get rid of risk by purchasing pre-customized software off-the-shelf, where the risk of doing the customization has been taken by a software developer (including time risk and acquisition cost risk - ie, having the entire group go together and split the costs of development is not a viable replacement).
In other words, I see proprietary derivates of free software as creating a social benefit. I see it as a social activity.
Anyway, What is scarce in a free software economy is developer time. To increase freedom through free software, we need to increase choice - which means to produce more software, with free software having a higher value than proprietary software. However, proprietary software has a higher value than no software at all, and proprietary software derived from free software has a large chance of feeding back to the free software it is derived for, thus adding value at two points. This means that the freedom to create proprietary derivates creates values in two forms: Both the software (and thus utility to users) created purely proprietary, and the developer time it gives back to the free software original. In my experience, the deliveries back are substantial - for instance, both the SCSI subsystem and the netgraph subsystem of BSD was given back by proprietary derivates. When I did a proprietary derivate of FreeBSD (delivering an ISDN router/web proxy/UUCP mailer to customers), about 90% of the changes we did were contributed back. The last 10% - and the possibility of keeping other changes if we had needed to - worked as pay for the rest, and would only have been relevant for a direct competitor (as they munged FreeBSD as a general system - I messed up the tty code to make other things work).
The GPL, by contrast, gets extra developer time the following ways (and that's all I have found - I would be glad to hear of more):
Eivind, who has nothing against guns per se, but think the way people see guns in the US (including "self defence" with a gun as an idea) is the most likely source of all the horrible gun violence in the US.
Eivind.
If the purpose of modern copyright is to encourage the creation of artistic works, I can guaran-fucking-tee you that taking away people's own creative work is pretty much the worst possible way to achieve that. Since granting the copy privileges make it impossible for other people to create the creative works they want (based on the first work), it's clear that giving the privileges is discouraging in some situations.
What is happening isn't taking away the work, anyway. We're saying that the originator doesn't have infinite privileges *when (s)he chooses to make the work public*. The moment somebody make a work public, it is only partially their work - because part of the value of the work is the fact that the work is in the minds of others (AKA a part of culture). Thus part of the value of the work belongs to those people. If too large privileges are granted to the originator of the work, then the trade goes bad.
Eivind.
This assumes that all elements of life in this reality resolve down to questions of evolutionary theory, which I think is false. --I tend to think that we are not living in a closed system; that there are a LOT of outside forces at work which dramatically affect the human species and which have little to do with natural selection, --that and the rules which govern our reality are infinitely more complex than is currently understood. I find this very overbearing. You're just listing vague claims and attacking people with them, using Ericssonian language tricks to bring about a hypnotic trance and a vague sense of agreement because you allow people to fill in meaning for themselves.
Dirty arguing. Make some concrete, testable claims. SAY something, something REAL. Not just something that use language to manipulate. Manipulative language is a dirty game, and some of us happen to know enough to spot it.
If you're not playing this game on purpose, you've probably hung around too many "alternative" groups and got it by assimilation - many people there have it naturally, and others increase the force of it through NLP training.
Eivind.
Rats' dreams are no doubt similar to our own, which is why torturing them and using them as 'things' is wrong... Don't worry, due to the innocence issue they've switched to using lawyers.
Eivind.
Eivind.
The difference in consumer voltage between Europe, Japan and the US is a non-issue - we transport electricity at a much higher voltage, and then transform it down close to the point of use. The same isn't quite true for frequency - it is synced at 50/60Hz in the grid - but there are production facilities in operation that produce it at a different frequency and convert it to the grid frequency using a frequency changer. You can read more about in Wikipedia's utility frequency article.
The main problem with interconnecting the continents is the power loss associated with long distance transmission. As far as I understand, this makes interconnection impractical at the moment - local storage (as in the reservoirs described above) being more economical. Superconductors may some day change this.
Eivind.
Of course, like everything else, it is a tradeoff. I have a fair amount of tolerance for people's behavior, and basically feel that all behavior is OK as long as it doesn't hurt people. However, I have fairly little tolerance for people hurting others because "they are brilliant and should be allowed to act that way". It is possible to both be brilliant and learn to treat others well - and I feel that the latter is reasonable to ask.
Eivind.
On a side note, I tend to be skeptical of all negative-to-radiation research with Swedish researchers, because my impression is that they have a history of debunked claims, sometimes with gross method errors, sometimes with minor ones, and usually with a bunch of publicity when they come with the claims, and little publicity when the claims are debunked.
Eivind.
Eivind.
Eivind.
Eivind, who would like GPLites to go RATIONAL, rather than go defensive.
I personally do not use the GPL, and have had the advantage of being able to reuse the codebase I know (FreeBSD) for a kernel-tuned ISDN-router/Squid proxy/UUCP mail system, and thereby be able to spend more time on it and contribute more code back to it. With the GPL, the projects I did most likely just wouldn't have gotten done, so no derivate - and less contribution.
Eivind.