The FreeBSD Project posted its periodic status report and the list of new features in the upcoming FreeBSD-7 branch to let people know where it is headed with operating system development. The project's goals have always been ambitious and technologically impressive, but it has not always followed through with their proper implementation. Will Linux users finally be able to switch to a cohesively-developed operating environment? This Linux user hopes so.
The primary difference between Linux and FreeBSD is that Linux is a piecemeal assembly of separately developed and governed projects, and FreeBSD is almost entirely developed as a single block of software. I would not want to be a Linux distribution provider -- imagine all the stress of trying to solidify a tested and stable release amidst the unpredictable, often late and off-schedule releases of the projects you depend on. If you wait too long to release, you'll be including software that is too old for many of your users' tastes. If you're unhappy with the bugginess of a particular Linux distro, the likely cause is the difficulty in compromising between releasing the latest software and having ample time to test it all together and fix interoperability and compatibility problems that arise. And what do you do about bugs in other projects? Linux distro companies and projects have a bad habit of dismissing those bugs, mumbling something about them being "upstream problems" or some other such nonsense.
The BSD operating systems, by contrast, frequently have nobody to blame but themselves. It makes a lot of sense to develop an operating system as a single project rather than a collection of other people's projects. The latter process is doomed to eventual failure, and can run into serious problems such as XFree86 licensing debacle. Not that FreeBSD is immune to those problems in all cases -- the XFree86 issue chief among them -- but it is in a better position to work around them.
So there are a lot of performance issues slated for inclusion in the next major FreeBSD release. That's kind of exciting I guess, but I wonder what the focus would be if people didn't publish benchmark results among operating systems. Sometimes I get the impression that software projects and companies take this Web publishing thing too seriously in all the wrong ways, and ignore all of the significant points that they should be listening to. Performance is not the big issue with FreeBSD; the issue that needs to be addressed is operating stability and the overall sense that this operating system works really well. There are still little problems with drive controller and network card drivers that either crash or slow down the system. Although it's improved greatly in the past three releases, I think this is still the biggest issue that FreeBSD faces. Once the small and varied stability bugs are cleared up, FreeBSD will become a superior alternative to messy Linux distributions both for server and desktop use. I can hardly wait to switch back to it.
Articles 
