One brutally simple way to categorise the difference between the BSD and GPL derived licenses is to say that the BSD people don't care what you do with their code as long as you're honest about where it came from, while the GPL people originally saw Adam Smith as the enemy and used licensing to fence out businesses which might want to rip-off community contributions.
Since then, of course, people have matured, things have changed, and the combined technical and managerial abilities of those involved with the FSF have made it a huge success -with the combination creating enormous internal strains only partially expressed in terms of increased, and largely defensive, stridency on the political side.
Their use of the term "copyleft" as they describe it below: is symptomatic - because it's really not a meaningful term at all, just their attempt to sidestep widely used terms whose political connotations they don't like: terms like copyright, trademark, patent, or intellectual property.
The two major categories of free software license are copyleft and non-copyleft . Copyleft licenses such as the GNU GPL insist that modified versions of the program must be free software as well. Non-copyleft licenses do not insist on this. We recommend copyleft, because it protects freedom for all users, but non-copylefted software can still be free software, and useful to the free software community.
That political agenda gets expressed, for example, pretty much throughout the FSF's discussion of about 32 GPL and over 100 non GPL compatible licences. Five examples, picked because the licenses discussed are widely used, illustrate this:
GNU General Public License, or GNU GPL for short.
This is a free software license, and a copyleft license. We recommend it for most software packages.
This is the original BSD license with the advertising clause and another clause removed. (It is also sometimes called the ?2-clause BSD license?.) It is a simple, permissive non-copyleft free software license, compatible with the GNU GPL.
If you want a simple, permissive non-copyleft free software license, the FreeBSD license is a reasonable choice. However, please don't call it a ?BSD? or ?BSD-style? license, because that is likely to cause confusion which could lead to use of the flawed original BSD license. ...
Common Development and Distribution License (CDDL)
This is a free software license which is not a strong copyleft; it has some complex restrictions that make it incompatible with the GNU GPL. It requires that all attribution notices be maintained, while the GPL only requires certain types of notices. Also, it terminates in retaliation for certain aggressive uses of patents. So, a module covered by the GPL and a module covered by the CDDL cannot legally be linked together. We urge you not to use the CDDL for this reason.
Also unfortunate in the CDDL is its use of the term "intellectual property".
Apache License, Version 2.0
This is a free software license but it is incompatible with the GPL. The Apache License is incompatible with the GPL because it has a specific requirement that is not in the GPL: it has certain patent termination cases that the GPL does not require. (We don't think those patent termination cases are inherently a bad idea, but nonetheless they are incompatible with the GNU GPL.)
Modified BSD license
(Note: on the preceding link, the modified BSD license is listed in the General section.)
This is the original BSD license, modified by removal of the advertising clause. It is a simple, permissive non-copyleft free software license, compatible with the GNU GPL.
If you want a simple, permissive non-copyleft free software license, the modified BSD license is a reasonable choice. However, it is risky to recommend use of ?the BSD license?, because confusion could easily occur and lead to use of the flawed original BSD license. To avoid this risk, you can suggest the X11 license instead. The X11 license and the revised BSD license are more or less equivalent.
This license is sometimes referred to as the 3-clause BSD license.
Notice that everything is either good (copyleft) or bad (non copyleft) - but review the actual licenses described here and it becomes obvious even to a non lawyer like me that the real difference between what Gnu thinks of as GPL compatible and non compatible licenses is respect for the financial value of intellectual property - a term Stallman explicitly attacks in the essay linked to from the CDDL summary above.
Two key paragraphs from that attack illustrate the emotional resonances driving the use of "copyleft:"
The term [Intellectual Property] carries a bias that is not hard to see: it suggests thinking about copyright, patents and trademarks by analogy with property rights for physical objects. (This analogy is at odds with the legal philosophies of copyright law, of patent law, and of trademark law, but only specialists know that.) These laws are in fact not much like physical property law, but use of this term leads legislators to change them to be more so. Since that is the change desired by the companies that exercise copyright, patent and trademark powers, the bias of "intellectual property" suits them.
The bias is enough reason to reject the term, and people have often asked me to propose some other name for the overall category -- or have proposed their own alternatives (often humorous). Suggestions include IMPs, for Imposed Monopoly Privileges, and GOLEMs, for Government-Originated Legally Enforced Monopolies. Some speak of "exclusive rights regimes", but referring to restrictions as "rights" is doublethink too.
And two more paragraphs from near the end give some good advice:
As a result, any opinions about "the issue of intellectual property" and any generalisations about this supposed category are almost surely foolish. If you think all those laws are one issue, you will tend to choose your opinions from a selection of sweeping overgeneralizations, none of which is any good.
If you want to think clearly about the issues raised by patents, or copyrights, or trademarks, the first step is to forget the idea of lumping them together, and treat them as separate topics. The second step is to reject the narrow perspectives and simplistic picture the term "intellectual property" suggests. Consider each of these issues separately, in its fullness, and you have a chance of considering them well.
I'm sure he's right about most of that, but unfortunately the over-simplification he warns of is also precisely the one GNU relies on to distinguish politically acceptable licenses from unacceptable ones. Basically, inventing a synonym for a term you don't like doesn't change its meaning; thus what the FSF says is that "the two major categories of free software license are copyleft and non-copyleft" but what they mean is that the two major categories are those which recognise a right to capitalize on intellectual property and those which don't.
What this reveals, I think, is a very deep conflict within the free software foundation's intellectual and emotional foundations. On the one hand, they're an idealistic organization whose work has generated enormous benefits to software developers and users everywhere. On the other they've failed in their primary political mission: to restrict the benefits to the community of contributors by taking capitalist motivations out of the picture.
Thus there are people who argue that Red Hat and Novel don't contradict everything the FSF stands for and that IBM's commitment of most of the money now driving Gnu/Linux doesn't amount to a takeover, but I think they're wrong. On the contrary it seems to me that the expression of IBM's corporate commitment to Linux through consulting and marketing represents the very definition of intellectual capital exploitation; that Red Hat's license is exactly what it seems: a proprietary license covering open source code; and, that Novel's patent agreement with Microsoft directly contravenes the spirit of the GPL.
Basically it seems to me that money corrupts, and money coupled with good intentions corrupts absolutely. Thus the FSF might have recovered its ideological purity after blinking for Red Hat precisely because Red Hat's intentions were broadly seen as good for the Linux community, but subsequently let itself slide into a far deeper pit by failing to respond to effectively when IBM used the GPLV2 rules to end run the FSF in support of the Linux System 390 port done at the Bvblingen skunk works.
What happened there was that the port extended proprietary hardware to make a proprietary system in exactly the way the GPL forbids the use of GPL'd code to extend proprietary software - something that's completely within the rules for hardware vendors and common among people who make, for example, custom handsets or Linux based embedded controllers.
So, bottom line, what? On the one hand nothing: IBM's actions were completely within the rights, obligations, and traditions associated with the GPL. But, on the other the entire GPL construct, because what this produced was an intensely proprietary system whose sole obvious purpose was to make money for IBM's shareholders by riding the wave of Linux support generated by the open source community and the myth of Torvalds.
And on balance? On balance I don't think anyone intentionally broke the letter of the licenses involved, but that the Bvblingen porting effort will go down in history as the straw that broke the FSF and correspondingly that the FSF needs to deal with these issues -probably in the end either by abandoning its political principles or by excising the Linux kernel from the FSF/GPL community umbrella.