Blogging Roller

David Berlind recently blogged about using JavaDB (aka Apache Derby) to provide the browser-based persistence needed for offline AJAX applications. The downside is that you need a JVM in the browser, but the upside is cross-browser portability. Oh, wait. Maybe the JVM requirement is an upside (for a second, I forgot I work for Sun).

Berlind didn't mention the mozStorage project, which is adding the open source SQLite database to Firefox. According to the mozStorage docs, you'll be able to issue SQL queries from JavaScript in the Firefox 3.0 timeframe. The downside is that it's Firefox only -- or may that's the upside.

We finally switched over to using the org.apache.roller package name in the Roller codebase. In related news: Apache license headers have been added to all source files and our release files have been cleansed of LGPL dependencies. Perhaps we'll graduate after all.

Over the weekend, I put my finishing touches on the (last) two new chapters for RSS and Atom in Action. Tomorrow they'll both be off to copy-editing, typesetting and then to the printers for publication in mid-June.

I really lucked out in the reviewer category. Thanks to Walter VonKoch of Microsoft's Windows RSS Platform team, who not only answered my questions but kindly offered to review the Windows RSS chapter. And thanks also to former co-workers Pat Chanezon and Alejandro Abdelnur, who reviewed the ROME chapter.

By the way Alejandro is back from Asia, blogging again and already coming up with cool new APIs for ROME. Checkout ROME.Mano, a pipeline framework for RSS and Atom feeds.

Looks like WolfBlogs is live and my alma mater N.C. State University is blogging with Roller. I don't know the specifics of the setup, but I believe they are running on a cluster of Solaris/SPARC servers and using a pretty standard Roller configuration (i.e. Tomcat and MySQL).

Via The Aquarium I see that Mark Hadley's work on Web Application Description Language (WADL) is now a Sun Technical Report. WADL provides a way to describe a REST based web application or service so that tools can discover services, generate proxies, etc. As I understand it, WADL is to REST as WSDL is to SOAP.

There's also something new since the last time I looked at WADL. Mark has added a section on the Atom protocol and examples that show how to use a WADL file to replace an Atom introspection document. Looks like good stuff to me. If you need an introspection doc for your REST based web service, why not use WADL?

Via Google, I found that there's also a WADL presentation on-line.

I've got  to admit, I'm itchin' to start using the JDK 5.0 features. I want to see how EJB3/Annotations stack up against the Hibernate/XDoclet setup I've been using for years now in various projects. And I'd love to check out Xzajo's new Hibernate plugin for Netbeans 5.5. But I can't do that at work because Roller is still stuck in JDK 1.4.2 land. I can development, test and deploy using JDK 5.0, but we're still keeping the Roller code-base 1.4.2 compatible. Sigh.

And that reminds me. Hibernate itself is becoming something of a plugin. Why? Because Hibernate is providing an implementation of the Java pesistence API.

For example, if Roller used the Java persistence API instead of the Hibernate API, we could eliminiate our Hibernate dependency and still allow Roller users to continue to use Hibernate. We'd be coding against an API with multiple implementations: the Hibernate EJB3 implementation under the LGPL, the Glassfish EJB3 implementation under the CDDL and someday the Geronimo EJB3 implementation under the Apache license. We could plug-out Hibernate and plug-in something else. Plus, we be waving good-bye to JDK 1.4.2 -- a very nice side-effect.

Neil Young is offering a free stream of his new album today only: Living With War. And while you wait for a connection/stream, check out Code Monkey (heard it on the MacCast).

Simon Phipps is asking for feedback on a white paper that explains Sun's view of open source licensing models. If you've been wondering why Sun favors different licenses for different projects, then it's good reading (despite the fact that he used my "gimme credit, gimme fixes, gimme it all!" line).

Ian Kallen: I'm not condemning AdSense per se, it's a great service and revitalized web advertising after the flame-out a few years ago. But black-hat SEO's have definitely cranked up the game over the last year or so and are putting the whole market at risk again. Google may not be particularly motivated to go out and find the abuse but they have to act against it when you bring it to their attention. Watch this video.

Yesterday, Microsoft released a second beta of IE7. I installed it, re-built and re-tested my example code. I haven't found any significant changes yet and none off my issues have been addressed. The item.id field is still an integer, unique only within the the feed. There is still no item.summary field. You can't parse feeds from disk. The Common Feed Extensions still live in the Simple List Extensions namespace. Oh well.

Andy Oliver: For all I know Google is search, email, maps and a large experiment where programmers who do cool things are bought, frozen and put in a crate after having all of their spinal fluid removed. Why spinal fluid? Well just like most of Google's betas it might be useful someday.

That cracked me up. You know what else cracked me up? Laser Cats killed on Saturday night. Not sure how well it will work on Tuesday morning.

This is definitely an exciting time to be at Sun. Solaris O/S and SPARC systems are back in a big way, the Java community is strong, comeback kid Netbeans is winning back developers and even the Sun Java App Server is catching a second wind. And now, we've got a new CEO who lives and breathes open source and blogging and participation and transparency and other things that matter to geeks like me. Financial results are not making the shareholder in me happy yet, but I get the feeling we're turning a corner and punching the accelerator. Congratulations to Jonathan and Scott on their new jobs.

Sean Montgomery discovers that JRoller.com doesn't fully support Unicode, but lots of other Roller-based sites do. I suspect that converting the old JRoller.com database over to UTF-8 is a non-trivial problem. Any MySQL folks want to lend the Javalobby a hand?

Microsoft has made freeware "express" editions of it's Visual Studio IDE products and SQL Server available free of cost. Supposedly, they were to become payware in November 2006. This is definitely a smart move on Microsoft's part; an effort to capture some of the armies of developers moving to free tools like Eclipse, Netbeans, MySQL, PostgreSQL, etc.

The free IDEs are stripped down versions of Visual Studio and with a few missing features. They're suitable for folks getting started with .NET 2.0 development using C#, VB.NET, J#, C++ and ASP.NET, but they're missing some key features you'll need for "real" development. The Visual Studio site has a feature-by-feature product comparison that shows the differences between the Express, Pro and Team System versions. Here's my summary of what's missing from the free IDEs:

• No source code control - you can't even hook-in SouceSafe
  
• No unit testing - you need the big-dollar Team System for that
• No XSLT support - but some XML tools are available
• No mobile device support - oh well (try the Netbeans Mobility Pack instead)
  
• No 64-bit compilers

I've tried a couple of the express editions and they didn't look very stripped down to me, except in comparison to the current crop Java IDEs. None of the features I was using in Visual Studio C# 2003 were missing.

I recently upgraded the RSS and Atom in Action examples to .NET 2.0 using the express editions of Visual Studio C# 2005 and Visual Web Developer 2005. My overall impressions were mixed. I was a little disappointed that I had to install two completely separate IDEs to get C# development ASP.NET development capabilties, but I was glad to see that Visual Web Dev. includes a test web server, so you don't have to muck with IIS. All and all, free is good and I'm definitely happy that my readers can build and run all of the examples in the book with free-of-cost IDEs from Sun, IBM and now Microsoft as well.

I'm happy to announce that my Roller talk was accepted for ApacheCon EU (June 2006). I'm slated for 10:30AM on day one (here's the full schedule). I'm giving basically the same talk as I gave at ApacheCon US, but I'll have to make some significant updates because Allen's been busy refactoring the Roller backend and the proposed Roller roadmap has changed pretty drastically since then (more about that  later). Here's the abstract:

Roller is the open source Java blog server that drives the popular Sun's employee blogs at blogs.sun.com, IBM developerWorks blogs, JRoller.com and numerous other blog sites. Currently making its way through the Apache incubation process, Roller is built on a host of Apache technologies including Struts, Velocity, Lucene, Jakarta Commons, XML-RPC and more. This overview, a primer for Roller users and contributors, covers the Roller feature set, architecture, lessons learned, project status and future plans.

Andi is already looking into travel arrangements. I think the stars might be aligned just right for a vacation in Ireland. The week after the conference is the Sun summer shutdown and the in-laws have bravely volunteered to watch the boys.

Triangle Citysearch: Rust-colored walls, wooden chairs and a colorful row of mounted lights give Cafe Cyclo a cozy atmosphere, perfect for quiet conversations. By day, expect a quiet lunch spot. By night, expect a jazzy social scene with understated energy and lots of charm. The menu is as loaded with drink concoctions as it is with munchies. This is the perfect place for a light meal; the Asian-influenced dishes range from Vietnamese spring rolls to seared tuna.

That's us: jazzy and social with understated energy. Come on down. Josh has the details.

I'd like to thank the folks who developed and run the FeedValidator, a valuable service that let's you know if your feeds validate against the RSS, Atom and commonly used extension specs. The warnings that it issues may be irritating and some can be safely ignored, but they're valuable just the same. I don't particularly like the warnings about <content:encoded> (which Roller now uses, by the way) and the style attribute, but I understand why they're necessary. If you want to whine about something, whine about the crappy RSS specs that we're all stuck with not the folks that are trying to help you understand them.

Update: In Roller RSS 2.0 feeds, we now use <atom:summary> for entry.summary (which is new) and <description> for entry.text (as we always have).

I'm a week late with this post. My first post was April 11, 2002 (also here) but the 17th is the date I remember because that's the day that the Roller article was published at OnJava.com. I started blogging with Userland Radio and moved to Roller in June of 2002 right around the time Leo was born. Ah, the memories...

Jim Hurley on the JINI-USERS mailing list: we feel that "Jini" would make a compelling top level project at Apache, and plan to make contacts there in hopes of successfully proposing "Jini" as a project.

Right on!

... an obligatory Sun-related post to counter all that borg talk

The news had already leaked out via blogs, but now it's official: Java Studio Enterprise is going open source as part of Netbeans 5.5. I'm really looking forward to 5.5. I'm most interesed in the Subversion support, but I'll definitely be checking out the 2-way UML modeler and the new XML tools as well. No doubt about it, the 5.5 feature set will lure a lot of new developers to Netbeans.