Amidst the laundry list of links I get sent every day to paw through, FOSS Licences Wars was a standout. Despite the word "wars" in the title, it's actually less vitriolic than such a name would suggest. It's one guy's take on the whats, whys and should-Is of the different open source licenses.
The author, Shlomi Fish, has been dealing with open source licenses of one kind or another for quite some time now. This document has been around since 2008 in various guises with constant additions and revisions. He leans strongly towards the public domain / MIT branch of licensing, and does not hide his dismay for the GPL due to its "political nature".
The fact that the GPL is as much a political manifesto as it is a software licensing methodology shouldn't surprise anyone. Shlomi's reason for rejecting it is that the politics of the license often force people to solve the same problem twice -- once inside the GPL and once outside of it. There is a good deal of software licensed under both the BSD-style and GPL-style licenses, he argues, which often have to be reimplemented from the latter to the former (but not in the reverse direction, because the BSD licenses aren't as restrictive).
Up until recently, most of the arguments I've heard in favor of the GPL pit it against closed source software generally, rather than how it stands up against other FOSS licenses. But the differences are becoming more important, not less, so it's crucial to have open debates (like this one) about what each license grants you.
Another issue Shlomi raises is the morality of bundling free and non-free material in the same package. To his mind it's not a question of the larger morality -- it's that data isn't software and shouldn't be lumped in with it:
I have issues with the entire Debian policy of including only free-as-in-speech material in their distribution, regardless of its type. I don't feel that non-software-content (e.g: documents in human languages, graphics, audio, video, animations, non-code binary data, etc.) should be treated by the same rules as software...
I myself am okay with this approach as long as there is some labeling to that effect. And I'd make an exception for documentation from the list he gives, since a program without its documents is that much less useful (and if you're smart, you'll make your program docs available in terms no less restrictive than the program itself).
No argument for or against a given license can be made without including at least some personal details about experiences with the license. That's the key thing Shlomi did with this piece, and I'd like to see others follow suit in the same detail.
InformationWeek has published an in-depth report on Sun's future under Oracle. Download the report here (registration required).