IMO

Someone recently pointed out a blog entry to me that referred to a project as having "languished." This is a project which has gone through, on average, a major redesign upgrade/refactoring every year for the past 5 years. Color me confused.

My server farms are almost exclusively BSD and Linux, and with the Linux ones, they are mostly of the Yellowdog variety (I have a lot of Mac h/w). When I get a new Intel system the debate should be "Install BSD or Linux" but it never happens. 95 times out of 100 I'll install FreeBSD. I know it's been hashed out numerous times, but I do wonder why BSD does not get more recognition than it does. Would it do better if there was a bigger "marketing push" from the various BSD communities? Or does the media simply do better with communities and projects on which they can attach a singular face to? It certainly appears that the latter is a major factor. A lot of the media has a hard time, for example, figuring out various ASF projects. They want to assign one person as *the* person behind a single project, and despite the best efforts of whoever it is they interview, that misperception leaks through. Of course, it goes without saying that it's a darn shame when a technically top-notch project and community such as the BSDs don't get the credit and recognition they deserve. I appreciate 'em and I want to thank them all for all they've done.

As I get older, I find myself much more careful as far as exposure to the sun. This is especially true when we take our annual beach vacation. I love spending time on the beach and in the water, but am almost paranoid about getting burned. When you need that vacation time to recharge, the last thing you need is to be uncomfortable with a nasty burn. Another great thing about beach vacations is that I can get a lot of good reading in. Some books I'll be bring this year include:

• Terrorism and Tyranny by James Bovard
• The Elegant Universe by Brian Greene
• Goldfinger by Ian Fleming (for like the millionth time)
• The Race by James Schefter

And, of course, my iPod will be bursting at the seams, because you always need tunes at the beach as well.

One key distinguishing feature of the ASF is that it places community before code. It is important that the community be heathly and active, since the logical result of that is a codebase which is active and healthy, and can continue far longer than any single developer's interest level. One way which we handle this is by avoiding the whole concept of "main" developer or "key" developer or anything else which places one developer above his/her peers. It's a concept which we've used since before there even *was* an ASF, and it's one that works for us. This works well since, for the most part, what the ASF does is create code which implements a known protocol or standard; there is an existing "framework" which the code then casts into being. I think there is one particular codebase class where this method isn't the best solution, and that is in the creation of new open source languages. IMO, for a language to be successful and useful, it should be internally and externally logically consistent. Naming constructs for example, should be easy to guess. This is easy when there is a very small subset of people (one or two) who define the language (after all, standardization usually happens after a language is widely used). And by restricting the decision-makers, it logically follows that the linguistic feel is constant and consistent. An example is PHP, which is a language I love and one which is on my short list of prefered languages. But even so, you have to admit the there is little logical consistency in how functions are named, argument ordering, etc. Compare this with a language that has a more controlling singular force, like Ruby or Python, and you can see that, in some cases, some open source projects do better with a benevolent dictator. Of course, there are exceptions, and this doesn't imply that languages-by-committee are any less capable than languages-by-a-few. But it is interesting nonetheless.

Over this last weekend, I retired from service my last 2 A/UX servers. These 2 machines performed secondary and tertiary backup services, for other more advanced servers, but I recall the day when all my Net/Web functionality was provided by Macs running A/UX. Those machines ran very well, and even the 2 recently retired servers still performed adequately, although they were quickly running out of steam, even for the drastically reduced workload assigned to them. A/UX was very good to me. First of all, it was the vehicle that really got be interested in the Net and in "open source," but it was also my association with A/UX that really got me my first notice on the Net. I still run into people who say "Are you the Jim Jagielski who was the owner of jagubox and the editor of the A/UX FAQ?" I learned a lot by porting various programs and applications to A/UX, lessons I still use today. It is amazing how a "unsuccessful" OS could have created such a long-lasting community. But even as unsuccessful as it might have been, you can see the heritage of OS X by looking at A/UX. When Apple dropped the m68k chipset and went with the PowerPC, A/UX was dropped as well. They were supposed to create the next generation of A/UX, called PowerOpen, but it never really happened. OS X is what A/UX could have turned into, and in a much quicker timeline, if Apple had decided that the Unix server-space was still a viable market. But we had to wait a long time for that perfect marriage of Unix and the Mac OS. For over 15 years, I've had at least 1 A/UX server running, performing its tasks with quirky efficiency. But no longer. It's earned its rest, and I'm only too grateful to let it have it.