Raible Designs

What's the best way to integrate Ajax into a Java webapp?

I received an e-mail over the weekend asking how to integrating Ajax into into the various web frameworks covered in my Java Web Framework Comparison Whitepaper. Below is my reply:

The best thing that I've seen is to use DWR, Prototype and Scriptaculous.
These will work with all web frameworks, and if you're using Spring on the backend -
DWR makes it easy to expose your beans as JavaScript objects.

Also, there's a number of tag library solutions that greatly simplify things:

  http://javawebparts.sf.net
  http://ajaxanywhere.sf.net
  http://ajaxtags.sf.net

I haven't used the first one, but I have used AjaxAnywhere and saw a demo of
AjaxTags from its developers.  They both look like they could be very useful.

For those of you using Ajax in your Java webapps - what's your advice? Do you use these same libraries or other ones?

This post was partially motivated by my desire to reiterate things that are so obvious.

Maven 2's Transitive Dependencies

I spent a few hours this week integrating Maven 2's Ant tasks into Equinox. Here's a things I discovered in the process.

• The JAR you need to include to use the Maven Tasks is quite large, weighing in at a hefty 833K. Regardless, it's easier to include a ~1MB JAR in CVS than it is 10MB of JARs. Equinox (w/o the extras directory) weighs in at 10.2MB before, 1.3MB after.
• The transitive dependencies feature needs to be cleaned up a fair bit. There are a *lot* of unnecessary dependencies downloaded for many libraries. For example, Hibernate downloads a bunch of JBoss JARs and the Display Tag downloads all the various web framework JARs. I found myself excluding almost as many dependencies as I added.
• Using in-line dependencies (where you specify each one in your build.xml) is cleaner and requires less XML (b/c of attributes vs. elements). I went this route at first and it worked well, but to allow greater flexibility (for those who might want to use Maven 2), I ended up creating a pom.xml. 48 lines in build.xml turned into 195 lines in pom.xml.
• The Maven Developer team provides fantastic support - both on the mailing list and IRC. Thanks Brett!
• Maven 2.0 RC was released today.
• I need to figure out a way to "install" the persistence/web framework options in Equinox. The easiest way from a maintenance perspective is to parse pom.xml to add new dependencies. Entity includes might make the project more extensible though - and easier to upgrade.
• Once you have pom.xml in place, it's pretty easy to make the build/test/package process M2 friendly.
• Maven 2 is faster than Ant.

Overall, it was a pretty frustrating experience, with lots of trail-and-error. In the end, I got it all figured out and the only "hack" I had to do was delete servlet-api-2.4.jar after copying dependencies into my WAR. This is b/c Ant is not smart enough to know that dependencies with a "provided" scope shouldn't be copied. I could probably have a 4th dependencies path to solve this, but deleting sounded easier at the time.

I'm not going to check anything in yet b/c I still have to modify all the installers to specify their dependencies. The main advantage I can see in using a download-dependencies feature is easier upgrading of dependencies. It'll also make a lot more sense in AppFuse b/c certain parts of project (i.e. appfuse-dao.jar) can be published as re-usable modules.

In the meantime, if you'd like to try out Equinox with this feature (works with both Maven 2 or Ant), you can integrate it into a CVS version with the following steps:

• In the equinox directory, create a lib directory and download maven-artifact-ant-2.0-dep.jar into it.
• Replace your build.xml with this build.xml (view diff against current).
• Download pom.xml and put it in the equinox directory.
• Delete web/WEB-INF/lib.

That's it! I'm sure some things can be improved - please let me know if you find any issues.

DWR and Script.aculo.us now included in AppFuse and Equinox

Last night, I integrated DWR into both AppFuse and Equinox. AppFuse has had Script.aculo.us integrated since 1.8.2 and it's been great to work with - so I added it to Equinox as well. I added these with the philosophy that it's easier to remove them than to add them.

Thanks to Joe Walker (DWR), Thomas Fuchs (Script.aculo.us) and Sam Stephenson (Prototype) for authoring (and supporting) these great open source projects.

Next up: integrating Ivy (or Maven 2's Ant Tasks) for downloading dependencies. Hopefully both will still allow bundling JARs in a release so both AppFuse and Equinox can stick with the 1-download-to-get-everything philosophy. The main reason I'd like to integrate a dependency downloading tool is to reduce AppFuse's size in CVS, as well as make it easier to upgrade dependencies.

Java in Action Presentations and OS Rot

I think I have a serious case of OS Rot on my PowerBook. Despite the fact that it's been extremely slow lately, I went ahead and used it to deliver my Comparing Web Frameworks talk at Java in Action. I was a little hesitant when I agreed to do this talk - mainly because it required me to stretch a one-hour presentation into a 3-hour presentation. I figured the best way to take up all that time would be to do some live coding. So I recorded a whole bunch of "Live Templates" in IDEA and went in to the talk thinking I could pull it off. To say the least, my Mac didn't cooperate and the "live coding" I did failed miserably. JR Boyens hits the nail on the head in his review. Cameron only attended the first part (before the live coding started) and it looks like I did pretty good in the first hour.

Lessons Learned: 1) Have a backup plan and 2) don't do Comparing Web Frameworks as a 3-hour tutorial. I've never had a backup plan in the 2 years I've been speaking at conferences. I've been pretty lucky though, my demos have always worked. I was due for a failure. For my afternoon session about Ajax and Spring, I moved all the live coding stuff to the Dell laptop I had with me. This worked much better, but I was again spat on by the Demo Gods and over half of my demos failed to work. Oh well, I guess it just wasn't my day.

The good news is that all the demos are available online. The master/detail applications I developed are already part of Equinox, and the Ajax demo is available at http://demo.raibledesigns.com/equinox-ajax. Features include an ajax-ified Display Tag with AjaxAnywhere, an editable text with script.aculo.us (click on a user's first name in the table), in page updates with DWR (on the detail screen) and a zip-code autocomplete/city-state auto-populate with script.aculo.us and DWR. If there's any interest, I can write up a tutorial on how each feature was constructed. In the meantime, you can download the equinox-ajax project from java.net.

After my Ajax talk, I was approached by a couple of the AjaxTags developers, and they showed me some very cool widgets they're working on. I definitely plan on digging into this project in the very near future.

Made it to Java in Action

I just arrived at Disney's fancy "Yacht Club" for the Java in Action show. Today was a fun day - Julie and I took Abbie and Jack to Magic Kingdom and had a great time. Abbie got to meet Mickey Mouse, Winnie the Pooh, Eor and Tigger. She was scared of Mickey, but warmed up to Pooh and friends pretty quickly. It was weird - it kept raining off and on throughout the day, but it didn't seem to put a damper on anyone's spirits. Unlike Colorado, rain doesn't cool anything down. In fact, the humidity seems to crank the temperature up a notch or two.

Tomorrow is a full day - I have a 3 hour tutorial on web frameworks in the morning, followed by an hour of Ajax + Spring in the afternoon. After that, it's back to vacation-mode until I return to Denver next Tuesday.

I'd post the slides from my talks, but they're starting to make less and less sense (in downloadable form) as I add more images and less bullets. Besides, I plan on coding and conversing for most of the talks. That's the fun part of speaking at conferences - who wants to listen to a presentation anyway? Why does the good conversation have to take place in the hallways? Can't it happen right in the room?

New Laptop

When I saw Russell's Why I Might Switch Back... post a couple of weeks ago, I found myself wanting to write a response. My response was going to be I completely agree and I was going to bitch about how slow my PowerBook is (once again). Then, later that day, I was doing something with iPhoto and I thought - I really do like OS X. It's the Mac hardware that I don't like. And it's not the look of the hardware (I love that), it's the fricken speed!! Most PowerBook users I know don't switch b/w computers a whole lot - whereas I spend 50% of my time on a fast Windows desktop. When I go from something that's so fast to something so slow, it's quite painful.

Last week, I started working with a new client - developing an application with Spring, Hibernate, WebLogic and Eclipse. Installing WebLogic on OS X was pretty easy, thanks to this article. Even remote debugging with Eclipse was pretty easy to setup. However, when I started running WebLogic locally and trying to debug it with Eclipse, it was extremely frustrating. I've never seen the spinning beach ball so much in one day. When other developers would watch me work, it was embarrassing how slow my computer was. And it's not like I had a whole lot running: Mail, Safari, Eclipse, WebLogic and iTerm.

Over the past couple of months, I've started debating if my next laptop should be a PC. It's not like I hate the Mac or don't like my PowerBook - but Java development on a Mac is far slower than on a PC equivalent. The problem is that I really like the PowerBook's form-factor. I'm so comfortable using the keyboard, right-clicking with the Control key, and all that jazz - that I'd probably have a hard time adjusting. I realize that a lot of my PowerBook bitching might seem unfair - as I'm often comparing a Desktop to a Laptop.

What I'd really like is two laptops: a PowerBook for doing all non-Java stuff and a PC for doing Java stuff.

My dreams came last Friday when my client handed me a brand new Dell Latitude D610. It's got Windows 2000, a 1.6 GHz CPU and 1 GB of RAM. To be honest, I expected it'd have a lot bigger processor. However, the fact that it doesn't makes it easier for me to show you how fricken slow my PowerBook is.

I used AppFuse for this test and ran ant clean war 3 times on each. I had ANT_OPTS set to -Xmx256m, JAVA_OPTS set to -Xmx512m and I'm using the latest 1.4.2 JDK available for each respective platform. It's possible my PowerBook suffers from some OS Rot, but it's still amazing how much faster the Windows laptop is.

• PowerBook: 58.3 seconds
• Latitude: 17.3 seconds

Holy ass-kicking batman!

My PowerBook has a 1.33 GHz CPU and 1 GB of RAM. It'd be interesting to do the see the numbers for a PowerBook with a 1.67 GHz processor. Here's to hoping OS X with a 1.6 GHz Intel processor can keep up with Windows for Java development.

Denver Tech Meetup on October 13th

From Stephen O'Grady:

For the Colorado residents in the audience, I'm pleased to inform you that I've finally scheduled the first of what will hopefully be many Denver Tech Meetups. The theory behind this is simple: there are meetups for just about every possible group out there, be it a company, database, development language or operating system, but nothing that I'm aware of that that brings folks from all of those different areas together, in one place. This is my attempt to rectify that particular problem, while having a few beers in the process.

If you'd like to attend, drop your email in a comment and I'll shoot over the Evite (you can join Larkin in laughing at me for using one of those), but otherwise it's Thursday, October the 13th, 6:30 PM at the Wazee Supper Club.

Bruce and I talked to Stephen about this a few months back. It's great to see the idea coming to life. Well done Stephen! I especially like the part about skipping the meeting and going directly to beers.

In related news, the Great American Beer Festival is in town this weekend. We're going tonight. 1,669 beers on tap is too good to pass up...

Hibernate Relationships with XDoclet Tutorial

I finally got around to finishing the Hibernate Relationships tutorial for AppFuse today. This initial version includes a howto for creating the UI with Struts. In the future, I'll add sections for creating the UI with Spring MVC, WebWork, Tapestry and JSF. Those should be easy now that the hard part is done. This is a first cut at this tutorial, so it's likely there's issues, bugs and things I did wrong.

Now it's your turn. If you have a chance, please try it out and let us know how we can improve it.

Jetty vs. Tomcat vs. Resin: A Performance Comparison

This morning, I did some comparisons between Jetty 5.1.5rc1, Tomcat 5.5.9 and Resin 3.0.14 (OS version). I ran AppFuse's "test-canoo" target, which tests all the JSPs using Canoo WebTest. I did this as a Servlet 2.4 application, and had to tweak some stuff in my web.xml to make it work on Jetty and Resin. Nothing big, just stuff that Tomcat let pass through and these servers didn't. One interesting thing to note that Resin requires you to use "http://java.sun.com/jstl/fmt" for JSTL's "fmt" tag URI, while Jetty and Tomcat require "http://java.sun.com/jstl/fmt_rt". This is with Resin's "fast-jstl" turned off - b/c everything blows up if it's turned on (I don't feel like coding my JSTL to Resin's standards, that's why I turn it off).

Below is a list of the average time it took to run "test-canoo" after I ran it once to compile all the JSPs.

• Jetty: 19 seconds
• Tomcat: 19 seconds
• Resin: 29 seconds

In addition, I tested how long it took for each server to startup - including the initialization of AppFuse.

• Jetty: 7 seconds
• Tomcat: 8 seconds
• Resin: 13 seconds

So what does all this mean? A number of things:

• I need to clean up AppFuse's web.xml a bit for 2.4 applications.
• Putting the database connection pool configuration in a Spring context file (vs. JNDI) makes AppFuse much more portable.
• Jetty isn't as fast as Jetty-lovers say it is (or maybe Tomcat just caught up).
• The open source version of Resin is much slower than the other open source servlet containers.
• I should restructure the build.xml to pull out Tomcat stuff and allow users to specify server deployment settings (i.e. in a ${servername}.xml file).
• Orion still doesn't support the Servlet 2.4 or JSP 2.0 specifications.

Should I change AppFuse's default web framework?

Currently, the default web framework in AppFuse is Struts. It's nothing fancy like Shale or Struts Ti, but rather Struts Classic. Even though Struts is not dead it's a pain in the ass to work with compared to other MVC frameworks like Spring MVC and WebWork. Yesterday, on the AppFuse Mailing List, I kicked off an informal poll about switching to a different default web framework. I think most of the people that choose Struts w/ AppFuse are choosing it b/c it's the default. Making another framework the default would likely same quite a few users a lot of headaches.

So which one should I make the default? Here's my thoughts from the mailing list thread:

I like Spring MVC and WebWork better than Struts, but I believe that WebWork is much easier to understand and develop with. Unfortunately, it's not well documented or marketed, so it's a bit difficult when you run into snags. Of course, if you contact the user community via forums or e-mail, answers flow quickly.
...
I'd like to use the framework that's simplest to understand. Right now, in my eyes, that's WebWork. I think JSF and Tapestry are excellent too (as component-based frameworks), but Tapestry's learning curve is difficult and JSF has a lot of issues (like everything is a post). Hopefully things will get better with JSF 1.2, but it's probably another 6 months before MyFaces supports 1.2 - let alone the app servers.
...
Maybe we should just drop Struts altogether - or replace it with Struts Ti? Unfortunately, it'll probably be a while before it's ready for production (I doubt it's that useable now).

Of course, if a WebWork Book was out - this move would be a lot easier. I did talk to Patrick Lightbody on IM yesterday and he said "it's done" and supposedly he has copies, but I haven't seen anything on the WebWork Blog to prove this.

A related question: how much would it hurt AppFuse if I dropped Struts altogether and went with something like Wicket instead? I'd like to keep that cap at 5 web frameworks. If I dropped Struts and added Wicket, I might lose potential users, which might not be a bad thing.