Last night, my friend Wilhelm Fitzpatrick gave the presentation last night at SeaJUG. At the post meeting beer and pizza, I asked him to sit down and convince me to get a PowerBook as my next laptop. I found this to be somewhat ironic, since I have convinced a large number of people to buy Macs over my lifetime. So I watched MacOS 9 and iPulse and the Dock, poked around in the file system, and measured the touch of the keyboard. I've been reading the Infoworld special on Apple, reminiscing of my Macintosh days, and as an Apple employee. I'm not quite there yet. Items that still concern me are:

1. Apple's lagging on JVM versions
2. Closed source application layers (Cocoa, etc)
3. Performance -- we'll see how G5 does on this
4. Cost of hardware, although in my case the competition is the Thinkpad T40p, which is comparable or more.

And then there's the other stuff:

• The editor in chief of ComputerWorld writes a scathing editorial about the problems with Blaster/Lovsan.
• Slammer crashes a nuclear power plant monitoring system.
• SoBig brings down train signaling systems at CSX.

To top it off, read this interview with Sterling Ball, whose company was shaken down by the Business Software Alliance, and subsequently dropped all their Microsoft software for Linux.

So why am I waffling? I must be crazy. We all must be.

Note: Due to pilot error, I overwrote the original version of this post, so I'm attempting to reconstruct it.

ComputerWorld and Wi-Fi Planet are reporting that the Washington State Ferries are going to trial Wi-Fi service on Bainbridge, Bremerton and Edmonds-Kingston runs.

One thing that I hope they do is extend the coverage to include the ferry terminals not just the boats themselves. Last night I missed (again) the 10:55 ferry after the post-SeaJUG beer and pizza. The next boat leaves at 12:15, so I was stuck in the Coleman Ferry Terminal with nothing to do. If they had Wi-Fi (and power jacks) in the terminal, I would have trundled my laptop along to get some work done. It would also help if the coverage extended into the waiting parking lots -- people spend a lot of time waiting for the boat.

The ComputerWorld article has some interesting details on how they plan to solve the backhaul problem.

Mick found a posting by Tom Lord, the author of arch, on some of the difficulties with subversion.

I looked at arch when it first came out because it was a pile of shell scripts, which put me off because I also use Windows, and I was biased towards subversion. But as the years go by, I can't help wondering why it is taking so long. It seems like the svn team is taking as long as the bitkeeper team took, but for something much less ambitious than bitkeeper -- no flames.

Surprisingly enough, the best version control system that I've used is IBM's CMVC, because it can manage changesets (althought it doesn't call them that), and it can tie them into a configuration management layer. I'd like to try aegis and a C or Python version of arch.

Over at the mod-pubsub blog, jp has a question: What is OpenDoc?

You don't need to take a red pill to get the answer to this question. OpenDoc was a cross-platform compound document architecture originally designed by Apple for the Macintosh and then morphed into a SOM using, Corba'ized, framework when IBM got into the picture. From the user-interface side, the key notion was the ability to build documents (applications) out of parts. The parts could communicate with each other and interact via scripting. Compound documents were also stored together but the storage was structured so that you could access the data stored by individual parts as well as the whole document.

The C2 Wiki has a pretty good description. Here's another one that helps show the vision.

The compound document metaphor is a perfect fit for microcontent (and other) applications. This is an old idea whose time has come (again). Unfortunately, there are so few people who remember what this was or how it worked, that it will be probably reinvented (and badly).

Mark Baker responded to yesterday's post with kind words. I'd seen Mark's posts on the FoRK mailing list, which I've been a lurker on for a long time.

He also posted some clarifications:

I don't dispute that the browser provides a relatively weak form of compound document framework when compared to OpenDoc and CommonPoint, but my emphasis at the time was in studying the architecture of the system to see if it prevented richer frameworks from being built by extension.

I think we're in vigorous agreemennt here. When I wrote the post, I had already accepted the notion of a universal front end, and was just trying to bring the various components into today's setting. As far as the compound document part of it goes, I think that systems like the ones that Mark identified, Dashboard and Spring would be easier to do an even more powerful, if we had a CDF that they could use. I'm not sure what the primorial but extensible OpenDoc is. I think that maybe Mark means the primordial but extensible universal front end. I know that we're making progress slowly, but it kills me to know that somewhere on a tape inside Apple and HP the CommonPoint and OpenDoc sources are slowly demagnetizing. Especially since I spent a fair amount of time evangelizing them, and also developing them. If IBM *really* wanted to help Linux, they'd give us the sources (and patent licenses) to these two products and let us rebuild them atop Linux and Mono.