Raible Designs

Ajax Framework Analysis Results

Way back in January, I wrote about how my colleagues and I were evaluating Ajax frameworks to build a SOFEA-style architecture. To make our choice, we used the following process:

1. Choose a short list of frameworks to prototype with.
2. Create an application prototype with each framework.
3. Document findings and create a matrix with important criteria.
4. Create presentation to summarize document.
5. Deliver document, presentation and recommendation.

When I wrote that entry, we had just finished step 2 and were starting step 3. I first wrote this blog post a week later, when we delivered step 5. Here is the comparison and conclusion sections of the analysis document we composed.

Framework Comparison
In order to evaluate the different frameworks against important criteria, we created a matrix with weights and ranks for each framework. This matrix shows how our weighting and rankings lead us to the winner for our project. You can view this matrix online or see below for a summary.

Note: Criteria whose values were identical across all candidates were weighted at zero. Charting capability was weighted at zero b/c we decided to use Flash for this.

This matrix indicates that GWT is the best candidate for our team to develop SOFEA-style applications with. In addition to the matrix, below are graphs that illustrate interesting (and possibly meaningless) statistics about each project.

Conclusion
After working with the various frameworks, we believe that all the frameworks were very good and could be used to write applications with. If all weights are equal, these frameworks were almost even when compared against our evaluation criteria. The graph below illustrates this.

Even after applying the weighted criteria, the evenness doesn't change a whole lot.

Without considering the even or weighted criteria, we believe the decision all comes down to what the developers on the project feel they will be most comfortable with. If you're developing with Dojo or YUI, chances are you're dressing up existing HTML and possibly using progressive enhancement to add more rich functionality. On the other hand, Ext JS and GWT are similar to Swing programming where you build the UI with code (JavaScript for Ext JS, Java for GWT).

The tools available for JavaScript development have gotten increasingly better in recent years. IntelliJ IDEA has a JavaScript Editor that provides many of the same features as its Java editor. Aptana Studio also has excellent support for authoring and debugging JavaScript. However, we believe the Java debugging and authoring support in IDEs is much better. Furthermore, we are more familiar with organizing code in Java projects and feel more comfortable in this development environment.

Based on this evaluation, we believe that GWT is the best framework for our team to develop SOFEA-style applications with.

Flash Forward to Today...
The core GWT library from Google doesn't have a whole lot of widgets, nor do they look good out-of-the-box. So early on, we experimented with two alternative implementations that continue to leverage GWT concepts and tools:

• GXT: a GWT version of Ext JS
• SmartGWT: a GWT version of SmartClient

Unfortunately, over the past few months, we've found that both of these implementations are too heavy for our requirements, mostly because of the file size of the generated JavaScript code. For example, a feature I wrote generated a 275K *.cache.html file using GXT. After determining that was too slow to give users the initial "pop", I re-wrote it without GXT. After a day, we had an application with *.cache.html files of 133K. Yes, that's over a 50% reduction in size!*

Because of these findings, we are proceeding with the core GWT library from Google and adding in new components as needed. It is cool to know you can make a UI "pop" with GWT, as long as you stick to the core - close-to-the-metal - components. For those applications that can afford an initial "loading..." state, I'd definitely recommend looking at GXT and SmartGWT.

* To make refactoring easier, I copied GXT MVC into our source tree and modified all imports.

Maven 2 Archetypes get a much needed improvement

Yesterday, a new version of the Maven Archetype Plugin was released. This release incorporates many of the improvements that were developed in a different project - code named "Archetype NG". The two major improvements are 1) you only have to use "mvn archetype:create" now and 2) you can create archetypes from existing projects.

I haven't tried #2, but #1 seems to work pretty well (especially since AppFuse archetypes are the first 9 ).

[Read More]

Don Brown Makes Maven 2 Not Suck

Don Brown spent some time over the weekend Making Maven 2 not suck:

While there are a few (very important, I might add) things Maven 2 gets right, there are a bunch that just suck, yet I use it at my day job (Atlassian) and in Open Source work, so in true Open Source tradition, rather than continue bitching, I'm doing something about it. I'm embarking on a quest to fix all the bits of Maven 2 that really annoy me and waste my time. I hope to get most, if not all, of the changes back into the codebase, but my personal deliverable is a build of Maven 2 that doesn't suck.

On his blog, Don lists a number of improvements he hopes to make. This weekend, he implemented the first three, which concentrates on speeding up remote repository access and downloading of artifacts.

First up, tasks #1-3. I implemented these changes in a bored Sunday afternoon and saw a example build (Struts 2 core) go from 3 minutes, 26 seconds to 2 minutes even, so a little over 40% performance improvement.

Interested, I decided to try Don's improvements on AppFuse. Since it fetches seemingly hundreds of artifacts from Maven's central repository, it seemed like a good testing ground. With a clean repository (rm -r ~/.m2/repository), a 8 MB/sec internet connection and "mvn -Dmaven.test.skip", I achieved the following results with the stock version of Maven 2.0.8:

[INFO] Total time: 7 minutes 40 seconds
[INFO] Finished at: Mon Jan 28 09:02:11 MST 2008
[INFO] Final Memory: 55M/508M

With Don's improved uber-jar, I received the following results:

[INFO] Total time: 5 minutes 17 seconds
[INFO] Finished at: Mon Jan 28 09:10:56 MST 2008
[INFO] Final Memory: 56M/508M

460 vs. 317 seconds = a 31.1% improvement -- Nice work Don!

When he implements #4 (Should support artifacts checked into the SCM in the lib/ directory so no external repository needed), I'll be a much happier Maven consumer. I've always wanted the ability to bundle all of AppFuse's dependencies for offline use like we did in 1.9.x.

Don - I'll buy you numerous beverages in Vegas if you add the ability to run a Maven command to put all a project's dependencies in its lib directory too.

Oslo in September

Wednesday we had a nice relaxing day in Oslo. After breakfast, we went on a 2-hour boat tour of the local fjord. The boat lulled me to sleep for most of the trip, but my mom kept elbowing me to keep me awake so I did see manage to see some things. Many islands were so small they could only fit a single house. They all looked like very enjoyable locations to spend the summer months.

After the boat trip, we browsed through an outdoor exhibit by Steve Bloom. The exhibit had some amazing photographs of wildlife - I may have to buy some to fill up my empty walls. After that, we had a bottle of wine on the docks and then headed back to check in to our hotel.

At the hotel, I had my most expensive cocktail ever - paying $26 for a Whiskey and Coke. We had a very nice dinner on Wednesday night and retired fairly early. It's nice traveling with my parents because you get a good night's sleep every night.

On Thursday, I had my talk at 11:45. I was speaking in "Gate 1", which has a stage in the middle of the room - which audiences on both sides. It was a bit difficult to talk to one audience without turning your back on the other one, but I think I did OK. I told folks I would post a link to my slides after my talk. Rather than continually linking to the individual PDF, I've created a Presentations page that will list my presentations from various conferences and such. The direct link to my JavaZone talk is here. One of the highlights of the show (for me) was meeting Aslak Hellesøy. I've always admired Aslak for all his contributions to open source.

WARNING: Please remember these slides that are merely a backdrop for my presentation. During the talk, I do say that some things are innacurate or have been fixed with recent (or future) releases. The slides don't really mean anything without the accompanying talk.

After JavaZone, my parents and I went to see the Fram Museum and then to Vigeland Park. The sculptures at Vigeland were pretty incredible, especially since we were there while the Sun was setting.

This morning, we hopped out of bed, had some breakfast and headed for the airport. After making it all the way to the gate, our flight was cancelled at the last minute. The reason is mechanical and apparently parts need to be flown in from London. Currently, it's 2:30pm and they have no idea if it will be a 4-hour or 48-hour delay. They bussed us over to the nearby Clarion Hotel, fed us lunch and gave us rooms. There's no seats available on other flights, so everyone on our flight is pretty much stuck. Hopefully we'll get out of here sometime tomorrow.

Update: Looks like we'll be on a flight tomorrow morning at 7:00 AM. They've given me no indication that I'm confirmed through Denver, so tomorrow night could be spent in Newark.

Do we even need web frameworks?

On the Struts mailing list, Ted Husted writes:

IMHO, if we had today's JavaScript/CSS/HTML environment available to us in 1998, then JSP, ASP, JSF, PHP, and all the rest of it, would not exist. We did all these things to make up for the shortcomings of the client-side environment, and, since then, the key shortcomings have been addressed.

I believe Ted has a good point. However, today's Ajax development toolkits don't help those applications that require Section 508 compliance. Of course, if your toolkit embraces 508 compliance, I'd love to hear about it.

AppFuse 2.0 Status

We're working hard on AppFuse 2.0. The good news is things are coming along really nicely. There's quite a few developers working on the project now and they've been a tremendous help. If I knew the move to Maven 2 would've inspired so much help, I would've done it a long time ago! We've got most of the archetypes created and we just need to work on documentation for the 2.0 release M1 release. As far as the roadmap is concerned, we've finished most of the code for an M2 release, but we still need to do documentation. I'm hoping to release 2.0 M1 on Thursday of next week.

If you want to try it, you can checkout the Hello World with AppFuse 2.0 video. After that, feel free to take it for a test drive using the QuickStart Guide.

New features in AppFuse 2.0:

• Maven 2 Integration
• Upgraded WebWork to Struts 2
• JDK 5, Annotations, JSP 2.0, Servlet 2.4
• JPA Support
• Generic CRUD backend
• Full Eclipse, IDEA and NetBeans support
• Fast startup and no deploy with Maven Jetty Plugin
• Testable on multiple appservers with Cargo and profiles

Speaking of goodies, I uploaded a bunch of AppFuse desktop backgrounds to Flickr. Thanks to Max Hays of Timberline Group for creating these.

Only 26 days left until the AppFuse 2.0 Release Party! Location TBD.

AppFuse 2.0 Status

I managed to get most of AppFuse's code moved over to a Maven 2 structure this week. I have tests working in the hibernate/ibatis projects, as well as the service project, but not in the web tier projects yet. I do have everything compiling though.

[INFO] ------------------------------------------------------------------------
[INFO] Reactor Summary:
[INFO] ------------------------------------------------------------------------
[INFO] AppFuse ............................................... SUCCESS [2.228s]
[INFO] AppFuse Common Data Module ............................ SUCCESS [1.681s]
[INFO] AppFuse Hibernate Module .............................. SUCCESS [0.809s]
[INFO] AppFuse iBATIS Module ................................. SUCCESS [0.630s]
[INFO] AppFuse Data Modules .................................. SUCCESS [0.008s]
[INFO] AppFuse Service Module ................................ SUCCESS [0.696s]
[INFO] AppFuse Common Web Module ............................. SUCCESS [2.817s]
[INFO] AppFuse JSF Module .................................... SUCCESS [4.410s]
[INFO] AppFuse Spring MVC Module ............................. SUCCESS [4.038s]
[INFO] AppFuse Struts 2 Module ............................... SUCCESS [4.954s]
[INFO] AppFuse Tapestry Module ............................... SUCCESS [4.042s]
[INFO] AppFuse Web Modules ................................... SUCCESS [0.007s]
[INFO] ------------------------------------------------------------------------
[INFO] ------------------------------------------------------------------------
[INFO] BUILD SUCCESSFUL
[INFO] ------------------------------------------------------------------------
[INFO] Total time: 26 seconds
[INFO] Finished at: Fri Aug 18 15:06:54 MDT 2006
[INFO] Final Memory: 11M/26M
[INFO] ------------------------------------------------------------------------

If you'd like to review any of the work done thus far, I've uploaded my latest work to:

http://static.appfuse.org/downloads/appfuse2-20060817.zip

The major issue I've encountered to date can be reviewed on the Maven user mailing list. If you happen to try things out, the best thing to do is post feedback to the dev list. I'm considering another device-free weekend, so I may not respond until Monday.

Comparing Open Source Application Servers

With all the recent hubbub about GlassFish, I decided to do a quick performance test this morning. I downloaded all the most recent versions of the various open source application servers, deployed AppFuse 1.9.3 (Struts version) on them, and ran "ant test-canoo" to see if any of them were faster than the other. This was by no means a scientific, isolated test. It also didn't take into account any performance tuning you should do on these servers, I just used the out-of-the-box settings.

I ran these tests on my MacBook Pro (2.16 GHz Intel Core Duo, 2 GB DDR2 SDRAM) with my JAVA_OPTS set to:

-Xms768M -Xmx768M -XX:MaxPermSize=512m -Djava.awt.headless=true

When typing "java -version" at the command line, I got:

java version "1.5.0_06"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0_06-103)
Java HotSpot(TM) Client VM (build 1.5.0_06-57, mixed mode, sharing)

Servers tested (in no particular order):

• JBoss 4.0.4
• GlassFish B48
• JOnAS 4.7.6
• Resin 3.0.21
• Geronimo 1.1

I'm pleased to note that all servers allowed me to deploy appfuse.war without using a console or command-line tool. They all support dropping the WAR in some sort of auto-deploy directory. Very cool! Secondly, I was able to successfully deploy AppFuse on all of them with no changes to AppFuse nor the server. Quite impressive.

My test consisted of the following:

• Copying appfuse.war into the appropriate directory
• Starting the server
• Running "ant test-canoo" from my $APPFUSE_HOME directory once
• Running "ant test-canoo" 3 times, recording the numbers for each run

Here's what I found:

Since I know you're going to ask about Jetty and Tomcat (the two main servlet-only containers), I ran the numbers on those too. First off, I tried Jetty 6 RC0. No dice - I got the following error when trying to start the server.

java.lang.IllegalStateException: Cannot initialize context because there is already a root application 
context present - check whether you have multiple ContextLoader* definitions in your web.xml!
        at org.springframework.web.context.ContextLoader.initWebApplicationContext(ContextLoader.java:173)

Since AppFuse deploys on all the above app servers, as well as Jetty 5.1.x, I'll chalk this up to a bug in Jetty 6. I used Jetty 5.1.11 for this test because I already had it installed on my machine.

I don't know that these numbers mean anything, but it was a fun experiment. For those of you who think these numbers might mean something, here's the rankings:

1. Tomcat 5.5.17
2. Resin 3.0.21
3. JBoss 4.0.4
4. Jetty 5.1.11/GlassFish B48 (tie)
5. Geronimo 1.1
6. JOnAS 4.7.6

Of course, the better test would be hammering each server with 1000 concurrent users (or a number higher than that) and comparing how each server holds up.

Better client-side validation with Prototype

I'd love to see something like Really easy field validation with Prototype integrated into Commons Validator and other web frameworks' validation engine. The only hard part would be looping through form elements and adding classes with the validation information. Other than that, it doesn't seem like it'd be too hard to implement. Hat tip to Ajaxian.com.

Carpal Tunnel

Every month or so, after working a long-ass week, my left arm usually starts hurting as if I have carpal tunnel. Usually, I go get a massage and it feels better the next day, or a few days shortly after. The Massage Therapist always asks me if I've been diagnosed with carpal tunnel, to which I reply "No." They also ask me if I have tingling in my hands or forearms, and I always tell them "No, I just get a dull pain in my forearms when I work a long week." So I've never really had carpal tunnel AFAIK, just symptoms every month or so.

That all changed this week. I started noticing the dull pain in my left forearm at TSSJS, and I started noticing the tingling in my left hand yesterday. I've never had tingling before. Furthermore, last week was a pretty light typing week (but I might've played cards too much ). So now I'm worried; I'll probably get a massage this week and I have a chiropractor appointment next week. The only think I can think of that might be causing the tingling is: 1) riding my bike to work, or 2) the cheap-ass crappy keyboard I have at work.

I went to the Apple Store and CompUSA to get an ergonomic keyboard tonight, but had no luck. The Apple Store only sell the standard Apple keyboard and CompUSA only sells black Microsoft keyboards. The M$ keyboard's will work, but it seems wrong to hookup a Microsoft keyboard to a MacBook Pro with a cinema display. I have a meeting in South Denver tomorrow morning, so I'm going to stop by Micro Center. Hopefully they'll have something good.

Carpal Tunnel is a scary thing as a programming professional. It's one of the few things that can put you out of commission as a programmer. It looks like I'd better start taking it seriously if I want to keep slingin' code for the next 10 years.

Related: Carpal Tunnel in May 2004.

A Week Later: I went to a repetitive motion specialist yesterday. They said that hand surgeons hate them b/c they can solve most issues. They worked my left arm and hand, and expect everything to be better with a couple more treatments. It already feels a lot better, but I'm also doing stretches every hour - which helps a lot too.