OSNews

Whats your rational for that conclusion? I do not see the correlation between license and development activity.

I would argue that the BSD license is, in general, more friendly to business. I also think its a much more liberal license, but thats neither here nor there.

I don't think the license has anything to do with the momentum that it gained.

Apache has a liberal license and is also widely used.

Linux got some major corporate endorsements early on which made it safe for other corporations to use. CTOs are wimps when it comes to adopting new technologies.

I also think the boost in popularity also had to do with Linux having a nice story and figurehead. People like it when an organization has someone at the top that they can relate to. FreeBSD has always been a quiet and anonymous group project.

I really like FreeBSD but Linux is on easy street when it comes to funding, well on the server anyways.

Nothing to do with the license as long as it's open-source. Personally I left FreeBSD because of poor Flash support, I know it's not FreeBSD's fault but I need excellent Flash support for my work, unfortunately.

Another good reason to use FreeBSD I don't want admins watching videos on the server - not that you can run flash with out X and why would I want that?

I don't think it's the license. I think Liquidator hit something. I don't use FreeBSD like I want because of the lack of real flash support.

Now the question is what should the FreeBSD guys do to provide what is important to you, i.e. Flash support? I think your answer would be useful if it could get to them.

Yes therefore I have to use Windows to some degree, because Flash in Linux isn't that much better as in FreeBSD.

Well I'm pretty much on that track myself. While corporations most likely love using bsd, mit, licenced code due to the few conditions that comes along with them, there's another thing when a company are about to contribute code. Giving code to prospective competitors whom may enhance that code and use it to compete against you must have a bad ring to it in the boardroom. However, contributing code under a licence that means that your competitors will have to do the same is obviously an easier sell.

In my opinion, that GPL is by far the most popular open source licence is not because so many programmers out there thinks proprietary code is evil, but rather that GPL's tit for tat mechanism makes perfect sense for collaborative projects. If you have no real interest in collaborative development then obviously BSD/MIT are a perfect fit and are often used for such projects like Google with VP8, V8 Javascript engine, Go etc, they are not really interested in people helping them develop this since they prefer to do it themselves (they DO have the resources after all).

Looking past corporations, GPL is attractive for spare time developers aswell, since it means that if anyone enhances their code they will get access those enhancements as the rights given to them as an end user. It also seems to be the licence of choice for software companies which chooses to open source their apps/games after they have left the commercial market (ID Software comes directly to mind and they certainly do not think proprietary code is evil).

Again, I'd say that the VAST majority of people choose GPL for practical purposes (not counting corporations, which OBVIOUSLY does) and not for ideological reasons.

In a perfect world all source code would be open, heck even unlicenced because noone would try to screw someone else over and everyone would be appropriately attributed and/or compensated, but in the world we live in I'd say GPL is generally a more attractive licence to contribute through, than the likes of MIT,BSD.

Yeah, GPL makes it safe for companies to contribute without worrying about competitors taking advantage without giving back. The tit-for-tat puts some restriction on what companies can do but it applies to all and thus enables collaboration. In a sense, it works by giving advantage to the group of companies which are collaborating against the ones which don't.

There would be a lot of other factors but one of the reasons why the effect is so prominent in OS development is that it takes inordinate amount of resource to develop and maintain modern OSes. If a group of company is collaborating on single project, it becomes extremely costly for isolated companies to maintain the same level of competitiveness, and as a result we've been seeing extensive consolidation of OS development effort centered around linux.

More liberal licenses sure work well for some projects but they're usually much more focused smaller projects which don't necessarily require similar level of consolidated effort.

It's really about the economy of things and that's the natural and beautiful part.

Surely explains why so many big OSS projects does NOT use the (L)GPL...
Apache and pretty much anything related to it, rubyonrails, python, perl, php, X.org, erlang etc etc.
It would certainly seem like not using the GPL is not a big factor for getting big companies to contribute to your project.

Contrary to other opinions I agree with you. What I also see as a big talent drain is the Linuxulator. They should have dropped the Linux emulator layer. There is a lot of OSS that is not closed and has value. Finally I believe they should focus on bringing new tech inside. Apple has published a lot of code and unfortunately the FreeBSD project did nothing to incorporate it. IOKit is a very big tool along with older XFree/IOKit bindings. I also believe that they should focus more on uKernel designs, revamp DDE-Freebsd and embrace Clang. At least they could spin off a BSD-NG design with all the above fresh tech and co-operate closely with Haiku and IllumOS.

LLVM/clang support is in the works, since they can only ship an ancient version of GPLv2 gcc in their base. Because of this mplayer-svn does not compile.

The other things you asked of would need significant manpower, and would be better served by forking and rewriting everything. Rewriting everything tends to create a new set of bugs though. The first few iterations would suck.