daily weblogs

books

reference

java study group

web design

webinfo (links)

about

adventure

recipes

photos

about us

Blog Attempted Murder

Last week, in a story I didn't really follow blow by blow, retired software developer and blog writer Dave Winer shot his very own beloved reputation in the foot. In a calculated yet characteristically incoherent manner, he pulled the plug on over 3,000 free weblog websites, in which people lacking the computer savvy or money resources to do so themselves, can write to their heart's content about whatever they want, as long as it isn't Dave.

At first it must have seemed like a generous move to him, hosting those sites for free. Magnanimous! But when the going got tough, Dave got going. That, is, he fled, or, as we used to say back in the sixties, "split". Yes, Dave split. Or at least he tried to. And then, to Dave's great discredit, he pretended that he hadn't done what he did, nor said what he said. It was all just an "outage."

Sorry, Dave. I don't buy it.

Here's what I think. And since it's just what I think, I'll make no attempt to prove it, these are just my conclusions about the way what Dave did shook down. I'm just a citizen drawing my own conclusions. I'm a person, not a company, and I draw conclusions based on the facts as I saw them happen.

For whatever reason, Dave decided to pull the plug on 3,000 sites. Then, when he got caught, and pilloried, and ganged up on by the ladies, he said to himself (but to nobody else) "Oops!" and changed his mind. But he refused to admit it.

Dave's health is not very good, he probably won't live all that much longer. But his reputation, his charisma, his honesty, his character? To me, that all died last week anyway. Goodbye, Dave.

p.s. no comments on this blog. As Dave used to say before the music died, get your own blog. Oops! Make sure it really IS your own.

Gavin King and Marc Fleury were kind enough, in a quick 2-hour session, to describe the technical underpinnings and the JBOSS business model for the future of Java middleware. King went first, describing the Hibernate object-relational mapping software, and then Fleury talked about aspect oriented programming in middleware, and how JSR 175 annotations will to a large extent replace XML configuration files for EJB 3.0, which will require Java 1.5.

There were several people there from JBoss, but none of them had a copy of gavin King's book on Hibernate, Hibernate in Action [buy at amazon] , for the simple reason that it has not yet been published. The chapters have been appearing one by one on theserverside.com, and you can buy an electronic version from the Manning early access progratm (MEAP) if your really have to have the hibernate book right now.

A lot of people use JBoss as an application server, because it is free. There is no "per-cpu" pricing, so it is perfect for large organizations that require a lot of servers to handle their incoming requests, like the travel industry. A lot of banks, a lot of telcos, and the army all use JBoss, since because the source is available it can be customized.

One of the ways it can be customized is through custom annotations, described in the JSR 170 document. This form of annotation is destined to replace nasty XML configuration files for enterprise java beans (EJBS), also known as XML Hell.

What Fleury has done is fascinating: he has hired the lead programmers of some of the top open-source projects to work for him, where they will continue to do the same thing as before: make internet programming simple and excellent, and open. And also make big bucks. And influence the future through standards bodies, with a little bit more clout than they may have had before. He has termed the business model for this "professional open source", but all it means is that his organization provides 24-7 support for all those open-source products.

It means something else, however: there's a strategic play too. The simplification of enterprise programming on the server side, using all the open-source products that JBoss understands so well, means that more people will want to approach it, fewer people can accomplish more in less time because a lot of XML configuration headaches will be largely eliminated. And because more organizations are using those softwares, the JBoss team that wrote them understands what they need and want, and can drive standards definitions for further simplication and what I guess I would have to call "architectural leverage."

Fleury was really impressed by the MacroMedia Flex architecture, which is what will be replacing Flash (or maybe is based on Flash). He also had some interesting comments on Aspect Oriented Programming (AOP, or simply AO), noting that in being a whole regime of providing callbacks, it was more than just a cut-point language. He also said that one Adrian Collier had gone through the Gang of Four book, and described all of the behaviour patterns as aspects, for example observer/observable. And this would have implications for MVC programming in the middleware world.

The java code that he showed was the first code I've seen at any recent lecture that actually looked like Plain Old Java Objects (POJO) code. It was pretty interesting, pretty exciting, and made me as excited about middleware as I've ever been (which is 'not very'). All in all, well worth attending.

To my general readers who have no idea what I'm talking about, I apologize.