Darwin's Theories - Java category

Seam Leaves Home (on good terms)

Tue, 25 Mar 2008 00:58:00 GMT

In addition to cranking out a new stable release (2.0.1.GA) and a new early access release (2.1.0.A1), The JBoss Seam framework has been renamed to "The Seam Framework" and has a new website, http://www.seamframework.org/. This represents a "growing up" of sorts, similar to e.g., Jakarta Tomcat becoming Apache Tomcat, and differentiates Seam from the many smaller projects that are also part of JBoss - see the new diagram at http://labs.jboss.org/projects/, and note that Seam is right there at the top of the web tier. While still funded by RedHat/JBoss, with this new domain Seam is obviously being positioned to compete head-on with similarly-named project The Spring Framework.

As part of "moving out", Seam has been given a new logo, see http://www.seamframework.org/Community/GetASeamIcon. This is used in their new web site, which is of course running on Seam and available as part of the downloads (http://www.seamframework.org/Download). Speaking of downloads, if you're using a version older than 2.0.1.GA, now would be a good time to download and upgrade.

Also, there is a new book about Seam, Seam In Action (http://www.amazon.com/gp/product/1933988401), which you can now buy in Early Access (http://www.manning.com/dallen/) - you get EA PDF's now and the final version when it's released, and optionally a paper copy then too.

Finally, JBoss Tools - the all-encompassing Eclipse tools for developing Seam and Hibernate projects - has been upgraded to 2.0.1.GA. You might want to download this as well.

RepRap and The End of Want

Sun, 09 Mar 2008 18:40:00 GMT

In the Star Trek: The Next Generation TV series, one of the most pervasive inventions was the "replicator", a device that could magically create any object, including a mug filled with "Earl Grey Tea, hot". This device ended poverty on planet Earth and on all planets that were allowed access to "Federation" technology. Need food? Just ask for what you want. The replicator has post-WiMax access to vast databases with the molecular formulae for every known food and drink, as well as every known object. It also apparently ended economics and economies, since the study of economics has as its deepest basis, its most secret heart of hearts, one person's need voluntarily to trade with another for something of value. If one can replicate anything one needs, why trade?

We are a years (if not centuries) away from being able to create such a replicator. But the first steps are being taken, and some of them with the same communal attitude that pervades ST:TNG's discussions of economics. RepRap is a project that is building simple "3D printer" replicators today. The project's web site has a subtitle of "Wealth without money..." These replicators can make small objects by a process known as fused deposition modelling (FDM for short). Commercial FDM machines cost around $30,000; the RepRap "Darwin" would cost about 1% of that, around $300-500 depending on how much you buy and how much you build.

The brainchild of Adrian Bowyer of the University of Bath in the UK, the RepRap project is not only building simple replicators, but giving away all the technology to make them, and even encouraging everybody who builds one to build the key parts for two more and pass them on. All the plans are freely downloadable, so you could build one without spending a nickle with the project. Though you'll find that their buying-in-bulk policy will save you money over buying some of the components directly. All the software, similarly, is freely downloadable from SourceForge. Written in Java so that it can run on almost any computer, the main driver software is covered by the GNU Public License which allows anyone to use it freely, and prevents it, or software based on it, from being re-sold in proprietary form. A far cry from patenting the ideas and charging license fees. Although, given today's idiotic patent system, I have encouraged Bowyer to patent any key innovations and either dedidate them to the public domain or assign them to a free-patent trust. But in fact the original idea for the replicator is neither from Bowyen nor from TNG. As the RepRap web site honestly points out, the general idea seems to have been invented by famed computer science pioneer John Van Neumann, under the name "Universal Constructor". I like "replicator" better.

To be really useful, this project needs a way of creating the computer files that describe objects. There are many CAD (Computer Aided Drawing) programs around that can be used, but this is time-consuming. What if you break a cup and want to replicate one to keep a set (relatively) intact? What you want for that is a "3D Scanner." Such things exist, at around the cost of a good computer server; see this 3D Scanner advertised at US$2,500 (web site is dysfunctional without some proprietary plugin). Presumably, the cost of these will come down, or mayhap some worthy RepRapper will publish a low-cost design for one that can be built on your existing RepRap replicator.

There are, of course, many other projects going down similar paths to openness. Open Source has been around for thirty years (long before Stallman). Open Hardware is clearly similar to early Ham Radio and do-it-yourself computers (there was even a home-made PBX). Asterisk is open-source software that lets you more easily turn a PC into a PBX. The OpenMoko cell phone project which I have written about here and here) has always made its software available under the GPL, and recently released the CAD files for all the mechanical (e.g., case and mounting) components of their cell phones.

Keep in mind, though, that today's replicator is to the ST:TNG replicator as, say, Leonardo da Vinci's glider is to the Space Shuttle. While TNG's replicator works at the molecular level, today's RepRap is purely an electro-mechanical "3D printer", recognizably a direct descendant of the "numerical-controlled machine tool" works of the 1960's and the inkjet printers of the 1980's. The first version (named after me, or more likely, one of my more famous distant relatives) can only make things out of thermal plastic. And a given object is likely to be all the same color. It cannot even deposit metal, thus, it cannot make circuit boards, and cannot be called "self-replicating", one test of Von Newumann's "Universal Constructor" (another key test is self-assembly). In short, RepRap could make the tea-cup (in about a quarter hour), but not the tea. In fact, it has made shot glasses that hold whiskey, though I haven't even started building a RepRap yet, so I don't know if the particular plastic used imparts any taste to the beverage. That said, it has made many of its own parts. And never forget that the US Space Shuttle is "recognizably a direct descendant" of da Vinci's glider. Nor, that a journey of a thousand miles begins with a single step.

"The Web Beans Manifesto" is good.

Tue, 27 Nov 2007 00:03:00 GMT

Gavin King is often thought of as the developer of Hibernate, but he deserves as much credit for making Seam, the web framework that unifies JSF and EJB3. Now he's going for the standards hat, making Seam into "Web Beans", a Java Community Process standard (JSR-299) that should get incorporated into future releases of Java EE. Check out this Web Beans Manifesto, which I agree with and which pretty much sums up why Seam/Web Beans, as it is now almost trite to say, "matters".

Mac OS X Users want Java 6

Wed, 07 Nov 2007 23:29:00 GMT

This would seem surreal to a mainframe programmer of 20 years ago. Bloggers are putting the string 13949712720901ForOSX into their Blogs so that a Google Blog Search/Count can be used to show that a lot of people want Apple to release Java 6 for Mac OS X. Bring it on!!

A Bigger, Better Tomcat Guide

Mon, 05 Nov 2007 01:40:46 GMT

My colleague in communications, Jason Brittain, has finished revising Tomcat: The Definitive Guide. The Second Edition was just released (October/November, 2007) and covers Tomcat 6.0. Coverage has been expanded in almost all areas; the first edition's 295 pages are dwarfed by the second's 462 pages. If you're using Tomcat, this is the book you should have to help you along!