Java

This forty-eight minute podcast presents a session from the EDUCAUSE 2007 Annual Conferece. The session, "Interactive Services on Mobile Devices for Higher Education," consists of a panel from the University of North Carolina at Wilmington discussing a new initiative to develop and test a wide range of information services accessible through mobile devices. This session compares the relative costs and advantages of applications based on SMS text messaging, WAP browsers, and Java MIDlets as well as the structuring of the relationship between the faculty start-up and the institution to develop and pursue the resulting initiatives. A PDF is also available for this session.

This session features:

Sun vice president and fellow Graham Hamilton has apparently left sun over their open sourcing of Java.

Hamilton, who in the past spearheaded Sun's work on compliance, interoperability and portability in Java sees the open sourcing of the platform as a bad thing because it erodes the technical uniformity of the platform, by allowing others to build incompatible versions. He is, of course, right. Open sourcing Java will allow third parties to fork the platform, but I have no doubt that the network effect would cause these to wither and die.

cheers, stuart

There's been huge coverage of the fact that Sun Microsystems has announced that it will release the core products which it has been marketing as Java as open source.

What much of the coverage doesn't mention is that what they're open sourcing is not "Java," since they're not relinquishing that trademark, but merely something technically indistinguishable from it. Sun is likely to provide a comprehensive test suite and allow those that pass with flying colours to use certain trademarks in restricted ways, much as they did a few years ago with their 100% Java program several years ago.

The move from Sun is a logical response to the increasing maturity of the already GPL'd classpath and gcj projects, which taken together amount to a reimplementation of Java. By GPL'ing the "real" Java, Sun prevents the open source community from abandoning it for the the new contender.

Having two separate implementations of Java is actually a great thing, because it promotes competition between them. The Classpath/gcj implementation, for example is already available on far more platforms than Sun's java, simply because it is built the same portable framework as gcc, the Gnu Compiler Collection.

I have written a Book Review for InfoQ on Agile Java.

This last semester I taught Java Web Development here at UNO. It was a great experience from the standpoint that it made me learn things even deeper. I hope my students got something out of it. I was also fortunate enough to have taken the approach of trying to facilitate learning rather than playing the "I know everything" game (in fact I don't :) ). This allowed me to leverage knowledge in the classroom and I think all had a better experience because of it. We went over many items and even put some of it into practice. The thing that really saved my rear though was Netbeans + Derby. Without that I probably would have created a bunch of people who decided that Java web development sucked. Even though it is only hard because it has gotten so complex and layered. Netbeans allowed me to skip all of the Tomcat setup. Derby allowed me to to skip the DB setup and Netbeans made integrating all of it a piece of cake. Great work guys! I am glad to see the Swing crew come though.

Sun have announced that their Sun Java System Portal Server (jsr168) system, is to be released as open source. They've already released some of the more minor components, a few portlets.

I'd like to put this down to high ideals on the part of Sun, but I find it hard. They've seen that the portal market is being consolidated by merges, both planned and in progress and they don't want to be left out in the cold. By open sourcing their portal server they place themselves to merge with the other contenders and greatly reduce the cost of supporting and maintaining the software going forward.

Bruce Eckel has written an article entitled "The departure of the hyper-enthusiasts" in which he suggests that a tipping point has been reached in programming ecosystem and that Java, long the darling child, is now on the way out. One of the new contenders is Ruby with it's Ruby on rails web framework.

In software engineering, ideas are largely embodied in formalisms and tools, and the most important of these are languages and their implementations respectively. Each generation of software engineering ideas leads to new languages and new implementations, with new ideas and a different balance of old ideas. Few of these languages take off in the real world, even if they are things of beauty to the academic software engineers, because in the real world the most important factor is retraining. There are huge numbers of technical people who have programming as part of their job, and retraining them to use a new language is a huge stumbling block, one traditionally solved by basing one language on a previous one. Just as Java was promoted as "better C++, with garbage collection," Ruby is "better Perl, with objects that work." Of course, promoting a new language by comparison to an existing one invites programmers to bring their old bad habits with them.

I have just returned from Java 5 catches up with C# by Barry Cornelius, a talk he gave here at the RTS. I've got to admit that while Barry is right, Java 5 is catching up with C# in the sheer number of features, I seem to disagree with him on whether this is a good thing.

This 30 minute recording with Alfred Essa, Executive Director of the .LRN Consortium, gathers his thoughts on open source, blogs, podcasts, java, .LRN and a range of other topics.

The OSS Watch Building Open Source Communities conference seemed to go pretty well yesterday. We had a broad range of people, with a broad range of interests and everyone seem to find something useful.

I had a great day, catching up with old friends and new, including: Gustav Delius University of York; Jim Farmer, Sakai Educational Partnership Program; Sean Keogh, OXILP; Bill Olivier, Development Director (Systems and Technology) JISC; Andrew Savory and Helen Sharp, Open University.