Linked by Eugenia Loli on Thu 23rd Jun 2005 18:11 UTC
Original OSNews Interviews One of my popular articles shortly after I joined OSNews in 2001 proved to be "the big *BSD interview" and so it is only appropriate to end my serving at OSNews with a similar theme. Today we are very happy to host a Q&A with well-known FreeBSD developers John Baldwin, Robert Watson and Scott Long. We discuss about FreeBSD 6 and its new features, the competition, TrustedBSD, Darwin etc.
Permalink for comment
To read all comments associated with this story, please click here.
RE: I think I get it now!
by Robert Watson on Fri 24th Jun 2005 16:16 UTC

> FreeBSD 5.x was a necessary step to get to 6.x,
> which promises to be t3h sn4ppy, and now with 5.4,
> it's really nice in it's own right. Just like we
> needed to get through the initial release series of
> GCC 3 and GCC 4 before we really reaped the benefits
> of what the teams have done.

Part of the problem with the development and maintenance of multi-million line source code bases is that some changes take half a decade or longer to complete. You can't do them outside the tree or you get left behind, and you can't stop doing releases for the duration, because then no one gets other necessary features and updates. The result is that they're rolled in in a staggered manner, with overlapping branches and releases. Similarly, old and new Apache versions overlap, old and new Linux kernel versions overlap, and even Windows versions overlap. This exact same pattern turns up in hardware -- the chip that's the bleeding edge computation cluster today will be in your toaster oven in a few years, but there's no reason to put it in the toaster oven today (if nothing else, it's a very expensive heating coil).

When you look at the FreeBSD front page and the release information, you'll see we are careful to mark which releases are "production", and which are "technology preview" or "developer snapshot" releases, so as to try and give a bit more information to the user about where it's appropriate to use various versions. Third party product developers will who are building embedded or derived products, such as firewalls or storage appliances, will often want to remain towards the head of the curve, because their product life cycles will often match our product life cycle. Large web hosts might track further back, as they'll have a less immediate need for the bleeding edge features.

One of the nice things about open source is that there's a lot of visibility into the intermediate steps in the development process -- you can see inside the product, how things are going, the bug report databases, etc. You don't get this openness with closed source products. Interestingly, this actually has a cost for open source in the enterprise: because people can see under the hood while it's being built, they assume that it's always in a state of being built. FreeBSD "-STABLE" production branches have quite long life spans during which they are *not* being built, and sometimes there are misunderstandings about when it's appropriate or not appropriate to grab a version. If you grab a developer snapshot, it will crash!

I think the world is becoming a lot more aware of the positive and negative sides to an entirely open development process, and so is able to take better advantage of it than they were a few years ago.