Skip to main content

Did Ye Get Healed?

Posted by editor on April 6, 2006 at 10:12 AM EDT

What not to try, catch, or throw

I don't know why, but I enjoy the heck out of anti-patterns. It's nice to see when someone recognizes bad ideas for what they are. Well, maybe not a bad idea, per se. If a pattern is a good way to handle a common problem, then an anti-pattern is a bad way to handle a common problem.

And don't tell me you don't see these anti-patterns all the time: spaghetti code, lasagna code, feature creep, etc. It's easy to think through failed projects, identify what went wrong, and know that it would be all too easy to make the exact same mistakes again. And that's the value of the anti-pattern: it shows how a given reaction to a problem may seem appealing, and how it has been shown not to work in practice.

Today's Feature Article, applies this to the process of dealing with exceptions, a constant point of frustration and contention among many Java developers. With so many best practices documents offering contradictory advice, maybe we can at least agree on what not to do. Tim McCune's Exception-Handling Antipatterns identifies a dozen worst practices for exception-handling, from the obviously-wretched "Catch and ignore" to the subtle side-effects of multi-line log messages.

In Also in Java Today, JBoss' Jason T. Greene relates a bizarre story of Taking RMH's JAX-WS Challenge. Replying to Richard Monson-Haefel's harsh evaluation of JAX-WS, Greene took up the rhetorical challenge to use JAX-WS to connect to a real-world web service such as Amazon or eBay. "So, in order to redeem JAX-WS, I decided to show just how easy it is by taking his challenge. I posted directions for a client that communicates with eBay in his blog comments. To be honest, I never expected to hear anything; however, to my surprise, he literally ate his words. I have to hand it to him for that. Although, he didn't have to go that far (I'm not that mean), a simple correction would have been nice."

Is nearly everyone adopting a service-oriented architecture for their enterprise project? According to the article Survey: 75% of Java Devs Looking at SOA, "more than three-fourths of enterprise Java devs are now working on (or plan to work on) SOA enterprise projects, according to a survey of 500 Java devs just released. The survey also found than some 30% of Java devs want to connect legacy C++ applications to their SOA infrastructure."

In Projects and Communities, Nigel Daley hopes to save you some CPU-time dollars with the blog Compute Server: locally debug your Sun Grid app. "This execution environment approximates the Compute Server execution environment on Sun Grid, with the exception that only one task will be executed at a time."

The latest episode of the GAME ON! Podcast features Sun Chief Gaming Officer Chris Melissinos discussing the recent Game Developers Conference, next-generation gaming consoles, and the Video Games Live concert at GDC, plus an interview with Jeff Kesselman, lead architect for Project Darkstar, Sun's new gaming server middleware project.

Tim Boudreau shows you how to Deliver your beans in a NetBeans module (the hard way) in today's Weblogs: "In Monday's blog, I announced a webstart-enabled standalone tool that will generate a NetBeans plug-in module that takes a JAR of Swing components, putting them on the component palette for users to drag and drop. Here I describe how to do the same thing using the plug-in development tools inside NetBeans."

In Product Architecture from a software perspective, John Reynolds writes: "I recently worked with an Enterprise Architecture team to develop a "holistic product architecture", but quite frankly we were not sure what "product architecture" meant. Here's what we came up with."

Davor Cengija has some advice on how to Install Instantiations RCP Developer on Linux: "Instantiations offers RCP Developer installer for any platform, but when you try to install it on Linux it fails. But you can install it on Windows and use it from Linux."

In today's Forums, ksak clears up some GlassFish naming conventions in the thread Re: Confused about java:comp/env: "java:comp/env is a special context available to Java EE components. It provides the abstraction that each web application and ejb component has its own private naming context. The 'java:comp/env' string is only used in the actual lookup() call. When defining the component dependency, it's up to the developer to pick a logical name that represents the dependency within its component naming context. There are recommendations for how to name things, e.g. prepending 'jdbc' or 'ejb' but those are only guidelines. The developer can choose any name he/she prefers."

Desktop developers are waiting to see what happens next in RE: databindings and org.jdesktop.binding.swingx: "You should be aware that databindings is the subject of JSR227. I suspect that there are quite a number of people looking forward to an announcement from that group - I for one have an application 'on the back boiler' waiting for some news - I'm reluctant to commit too much effort to the existing codebase since we are told that it could change quite significantly."

In today's java.net News Headlines :

Registered users can submit news items for the java.net News Page using our news submission form. All submissions go through an editorial review before being posted to the site. You can also subscribe to the java.net News RSS feed.

Current and upcoming Java Events :

Registered users can submit event listings for the java.net Events Page using our events submission form. All submissions go through an editorial review before being posted to the site.

Archives and Subscriptions: This blog is delivered weekdays as the Java Today RSS feed. Also, once this page is no longer featured as the front page of java.net it will be archived along with other past issues in the java.net Archive.



What not to try, catch, or throw
Blog Links >>