This is technically true, but I want to clarify this point using Redhat's distributions as example. Redhat is perfectly allowed to bundle GPL'ed software, write GPL'ed installers, test it all using GPL'ed testing software, then SELL that distribution with a NON-GPL license.

How is this possible? Well, the GPL says that anything directly derived from something that is GPL'ed must also be GPL'ed. The distribution isn't derived from anything GPL'ed ... it's built on top of things that are GPL'ed. That's the difference and it's an important difference. The distribution isn't selling the GPL'ed code, it's selling the organization of said GPL'ed code. In a very real way, it's taking a sysadmin's knowledge of how to put a Linux OS together from scratch and putting it on a CD. That isn't GPL'ed just because a sysadmin uses GPL'ed software.

Here's another item about the viral GPL - just because you use gcc to compile your program doesn't mean it must use the GPL (unless, of course, your program is a modification of gcc). I hope this clarifies things.

It is possible for something in the public domain to be ... patented.

IANAL, but I believe you're wrong on this one. Works in the ublic domain count as "prior art" and demonstrably publishing a work (where "work" is fuzzy, because for the most cases copyrightable works aren't patentable anyway, software in the U.S. being one notable exception) into the public domain prevent them from being patented by anyone else. A friend of mine used to work for a large manufacturing company in Germany and his job consisted of reviewing patentable works which someone else had already decided weren't worth patenting to see whether they were worth publishing so noone else could patent them. All your other points about copyright etc. are valid AFAIK.

IANAL, either, but it sounds like you're right.

Update: OK, I *thought* I had seen something about this. The Public Patent Foundation discusses this issue.

The GPL is viral. Its intent is to "infect" derivative works to prevent someone from trying to grab the rights to use and distribute the software (or other product).

I have never liked the term "viral" being associated with the GPL. I assume the term is used because a virus can be spread from one person (piece of software in this case?) to another. The GPL only "infects" its children (derivitive works). If we want to stick to biology then this is just heredity. Making the GPL sound like a disease is very misleading...

The GPL is viral.

Completely untrue. Copyright is viral.

OK, perhaps I need to clarify what I meant by "viral". I did not mean, "pernicious" or "evil". I meant that the license propagates itself through its descendants. Perhaps I ought to have used the word, "genetic" as a metaphor, rather than viral. In any event, I did not mean to suggest that GPL was in any way evil or wrong. I agree with you that copyright is the culprit, not GPL.

I'm not sure I explained my point fully. The concept of "derivative work" is the so-called viral property and that comes from copyright, not the GPL.

Not entirely. The various BSDs, even Apple's version, are free as in beer. Sometimes the intent may be that you have found something useful and think others might find it useful, too. (Which is basically the reason Larry made Perl available to the world, if I read the histories correctly.) And I'm sure ego-stroking is part of the intent, too.

Umm, how is software licensed under the BSD license not "free as in speech"? You can take BSD-licensed code and relicense it under a more restrictive license (e.g. the GPL or a proprietary license), but that doesn't take away from the fact that the original code is and remains Free/Libre Software.


All dogma is stupid.

Good point (I didn't know Apple's was free). No doubt there are many different reasons why people write the software they do - Larry wanted to give folks a useful product; Linus wanted a Unix clone; Stallman wanted some other goal, perhaps. And as you noted, probably getting ego strokes, recognition, was a factor for many. But that's the motivation for *writing* the software, not for choosing the license. It seems to me that, for whatever reason people wrote the software, they placed it under GPL (or similar) licenses to keep their creations 'free'. Why they wanted to keep it free may vary - some might just want to make sure they got credit for the work, which often doesn't happen with proprietary licenses. Some - the Mother Teresas of the IT world - might genuinely want to share their work with the world, so that everyone can have access to good programs. No doubt there are other motivations - heck, probably lots of people do it just to spite a certain large software company whose name I won't mention, but whose initials are "Microsoft".

Ultimately, only the people who actually did the licensing can say with certainty what their motives were - and they might not be altogether clear about it, either.

BTW - love the "Yet Another Perl Schlub"...

I can offer at least my own motivation for using open source licensing: I consider it unethical to treat ideas as "property". I don't use open source licenses to achieve an aim so much as simply because, given an option, I think it would be wrong of me to do otherwise.

print substr("Just another Perl hacker", 0, -2);

- apotheon CopyWrite Chad Perrin