The path ahead
by Greg Lehey

Last month I discussed BSD's place in the UNIX community, and compared it to politics: System V on the right, Linux on the left, and BSD in the centre. This comparison was not intended to be definitive, but rather to arouse thought and discussion. I didn't get much discussion, but Kirk McKusick came up with one comment, in an unrelated context: when asked about copyright and the GNU/Linux alternative ``copyleft'', he stated that BSD had something more like a ``copycenter'': the attitude is ``Take this, go down to the copy center, and make as many copies as you like''. My case rests.

The demise of BSD?

In these articles, I've always tried to talk about issues which apply to all BSDs. This month, I'll make an exception and talk primarily about FreeBSD. Before you accuse me of unfairness, read the article: it's about FreeBSD, but it has a number of messages for NetBSD and OpenBSD as well.

It's been over a year since Wes Peters talked me into sharing this column with him. Initially I wasn't overly confident that much would come of it: all we had were some people who were prepared to advance the FreeBSD cause, some more virulently than I liked. Even the idea of ``BSD'' advocacy, as opposed to FreeBSD advocacy, upset a number of people. Things have changed, and it seems time to look back to where we have come from, and to look on to where we think we're going.

It would be exaggerating things to say that the BSDs are now at a point where they could merge into a single project--as I have pointed out in the past, I don't even think that would be a good idea--but we're certainly a lot closer than we were two years ago. I do believe that Daemon News has been a significant contributor to that change. One small change that I particularly like is that we now talk about BSD; two years ago, it was either the individual flavour or ``*BSD'', a symbolism I personally dislike. Dropping the * doesn't just make it look better in print, it reflects a unity which just wasn't there before.

So what has really changed in the last two years? In the Real World, the big thing was that the public at large gradually came to realize that Microsoft wasn't the only way to go. The DoJ trial helped there, of course, but also the message came through that software didn't have to be so expensive, software didn't have to be so slow, software didn't have to be so buggy. Eric Raymond chose his time right and launched a press campaign about ``Open Source''. The press loved it, and various paranoid members of the BSD community predicted the imminent demise of BSD UNIX. Eric Raymond was not a popular person in the BSD world.

I think this was unfair to Eric. He had the gift of analyzing the movements in the Open Source community and being able to explain them to outsiders, in particular to people who wanted to deploy software, whether they be corporate IT managers or private software users. This is something that previous people had either not considered a worthwhile objective, or one which they had not been able to achieve. Linux has benefited enormously as a result. The only possible way that BSD could be seen to have suffered is by comparison with Linux.

Nevertheless, BSD also benefited indirectly, admittedly with some delay. In this connection, people who have been around for a while will recall that Eric has a BSD background. He's on record as saying that BSD has cleaner internals and a more elegant architecture than Linux. He's not out to play off one against the other.

One indication for the changes was the FreeBSDCon which took place in October 1999 (in Berkeley, of course). I think this is one of the most significant things which has happened to BSD since the disbandment of the CSRG. Sure, it's only FreeBSD, and it didn't cater for NetBSD or OpenBSD (or for BSD/OS, for that matter). But that can change. The important thing is that BSD has now become big enough to have its own conferences.

The FreeBSDCon

From my viewpoint, the Con, as it's called, was a logical successor to the meetings between FreeBSD people at the USENIX meetings in New Orleans LA in June 1998 and Monterey CA in June 1999. For many of us, these were the first times that we had met in person, and we found it an interesting experience.

Nevertheless, organizing a conference for the first time is a risky business. I've seen the results of over-optimistic conference plans several times in the last year, and Maddog Hall gave some further examples at the AUUG conference in Melbourne. Professional conference organizers are a necessity, but how do you find a good one? As a result of other incentives from the AUUG conference, I have been doing some investigation on that front, and the enormity of the problem has become particularly apparent. I don't think I would make a good conference organizer.

In addition, the big question was: who will come? Five years ago, the BSDs were very much hobby operating systems, the few commercial sites that operated with them notwithstanding. But hobbyists don't pay several hundred dollars for a conference, and even though we now have many more commercial users, they were an uncertain factor: after all, they were the ones who paid nothing for an operating system. Why should they pay money for a conference?

My fears were unfounded. The average daily attendance at the conference was about 330, well above expectations and about as many as the organizers could handle. The conference was as professionally organized as any I have attended. For the first time, I realized that the FreeBSD community isn't just the tightly-knit community that has grown up over the years: we have real users out there. Some of the more interesting things I learnt at the conference were:

And the other BSDs?

As I promised, this story is (almost) entirely about FreeBSD. So FreeBSD is going to win and NetBSD and OpenBSD are doomed to oblivion? No way. Remember the prophets of doom who said that BSD would be wiped out because of the popularity of Linux? Well, not only has that not happened, I strongly believe that, in some measure, we have the Linux advocates to thank for the publicity that has made FreeBSD as well-known as it is. One indicator of the direction things will go is that the publishers we have been talking to have shown interest in publishing NetBSD and OpenBSD books. In one case, OpenBSD has jumped the gun: O'Reilly has a book on firewalls which includes significant coverage of OpenBSD. So far, they haven't published anything on FreeBSD. In general, though, we can expect the route to be the same as in other matters: first, books will appear on FreeBSD, and later on NetBSD or OpenBSD. The authors and publishers of the new books are all open to coverage of all BSDs, but it makes it difficult to read if the books constantly say something like ``If you're running FreeBSD, do this. If you're running NetBSD, it's like that. On the other hand, with OpenBSD do the following''. The general feeling amongst the publishers is that there's probably enough volume for separate books (maybe sharing a common core) on all three operating systems.

BSDs of the world, unite!

Will we have combined BSD conferences in the futures? It's possible, but I wouldn't bet on it. It's interesting to see that each BSD seems to have its own very distinctive culture and conventions, and though open confrontations are largely a thing of the past, there's still a way to go before we all behave in the same manner, if indeed this is desirable. Certainly I'm sticking by my guns: I don't think it is an advantage for the FreeBSD, NetBSD or OpenBSD to merge with another. Each flavour derived from 386BSD for different reasons. Some reasons may have changed, but many are still valid, and the choice also helps encourage alternative implementations, which gives us a much better chance of coming up with an optimum. Now a common API--that's another thing. I'll rant about that some other time.

What will the future bring?

Where do we go from here? It's difficult to predict. It's my guess that things will gradually become more commercial, but there will be limits which will stop the overcommercialization of open source software, specifically: Whatever it is, we can be sure that it will be interesting. Hold on and enjoy the ride.