Raible Designs

David Geary won't learn Tapestry, but he'll write a book about Rails

Sorry David, but I have to call you out on this one. Yesterday, you wrote a long post about how you won't use Tapestry b/c you have bills to pay.

Do I use Tapestry? Heck no. I have a mortgage to pay. Besides, I'm so comfortable with JSF that I don't know if it'd be worth the investment for me to switch to Tapestry. Also, JSF already enjoys more industry support and that gap will widen considerably over the next couple years as Tapestry maintains a small but rabid group of followers in a niche market, whereas JSF will dethrone Struts as the reigning king of Java-based WAFs. JSF will eventually have support for HTML views and custom components devoid of Java code, in addition to many other cool features such as built-in AJAX support and client-side validation.

Today we find that you're writing a book on Rails. So rather than spending the time to learn Tapestry b/c it doesn't pay the bills - you're going to learn Rails? What makes you think it will pay the bills better than Tapestry? Is Shale still the next big thing for you - or do you just like writing books?

Tapestry is the best Java framework available today

David Geary:

So what's the best Java-based framework available today? It's a very close call, IMO, but I'd have to give the nod to Tapestry at the moment. I really like Tapestry's pure separation of HTML and components and the ability to create custom components without any Java code. That gives it an edge on JSF, which, like Tapestry is one of what I refer to as 3rd generation WAFs, that support components and a server-side event model.

He goes on to say that he'll likely continue to use JSF (with Shale) because it pays the bills and will dethrone Struts as the most popular - which will obviously lead to more gigs. I especially like this part of his post:

After I get client-side validation and file uploads added to Shale I want to turn my attention to Tiles integration, AJAX support and exploring Tapestry-like views that strictly separate HTML and component definitions. For me, those are the most exciting areas of Shale.

I agree that JSF definitely needs Tapestry-like HTML Templates. Shale definitely sounds cool, but I find it funny that it takes yet another framework to make JSF usable. Hopefully Shale will prove a lot of ideas worthwhile and end up as features in JSF 2.0.

[Microsoft] Day 2 Morning

Looks like I'm going to miss the good stuff today. There's a Company Store visit this afternoon at 5:00 - and they're giving us $120 in credit. Damn, apparently the games are pretty cheap there. They've also arranged for a Longhorn Demo during the company store visit. I did manage to get internet access with my phone and Ben's charger - so I should be able to do some real-time updating this morning.

I think one of the coolest things about this conversation is we're learning about how software is developed on a huge scale. These guys develop more software than anyone else and they do it on a very large scale. Can you imagine developing software for 90% of the computer users out there?! That would be nuts.

Oh boy - now we have an audience member ragging on the guy for Windows. He thinks it should be open source because OSes are going to be commodities soon. Personally, I don't think there's anything wrong with having a closed-source Windows. But I also don't see a problem with closed-source Java. What's wrong with companies making money? The main reason I'm in this industry is to make money - so what's wrong with the corporations doing the same thing?

.NET CLR Architecture

Started work on the Component Object Runtime (COR) back in '97. It was a small incubation project in the MTS group round metadata and compiler integration. Several API's still use the "Cor" prefix and engine DLLs are named MSCOR???.DLL. Apparently, this was all announced at the '98 PDC in Denver. Lots of code/marketecture names to follow: COM3, COM+ Runtime, NGWS Runtime, Universal Runtime (URT) and finally the Common Language Runtime. The big unveiling of CLR was at PDC 2000 in Orlando.

CLR Design Goals: Simplify development and deployment for classic Win32 programmers. Unify programming models, provide managed environment and support multiple languages.

The CRL is Language Neutral and has a commons set of features (i.e. Generics) guaranteed to be in all languages. Supports strongly typed languages, dynamic languages and functional languages. Because compilers are always targeting MSIL and the metadata, you get for free: shared object layouts and cross-language inheritance, exception handling, reflection, remoting and integrated tools for debugging and profiling. MS is very excited about having Jim Hugunin and IronPython on board and they expect to add even more killer features for dynamic language authors in the future. I wouldn't be suprised to see an IronRuby implementation in the next year or two.

CRL Version 2.0: More class libraries: collection classes, serial port, etc. Generics, 64-bit support (IA64, x64). RAD support: Edit and Continue, Just My Code, single-click deployments. SQL Server: fiber mode, integrated security, loading. Avalon, Indigo and Windows Longhorn.

Is this a boring session? Yes, but I'm sitting in the front row today, so it's a bit easier to pay attention and take notes. I'm doing AppFuse development for the most part and taking notes while AppGen tests are running in the background.

The CLR will be hostable in numerous environments: ASP.NET, VSTO for Office, DB/2 Stinger, Oracle 11g and SQL Server 2005. Oracle and DB2 are out-of-process with v1.1, SQL Server will be in-process with v2.0.

ASP.NET 2.0 by Scott Gu

• ASP.NET 2.0 Application Services APIs: Membership, Role Manager, Personalization, Site Navigation (XML and CMS-based), Database Caching, Health Monitoring. These are all based on a new "Provider Model Design Pattern" that seems to be a set of interfaces that you can easily switch or implement yourself. Out of the box providers: Windows, SQL Server, Access (ha!).
• ASP.NET 2.0 "Page Framework" Features: Master Pages (like Tiles, moreso than SiteMesh), Themes/Skins, Localization, Client-Scripting (using XMLHttpRequest). Holy shit - the guy said that they're testing their controls, particularly the scripting ones on IE, Safari and Firefox! That's pretty surprising to me.

VS 2005 has a new feature - where you don't need to have a web server installed, it'll work off the file-system. A new project doesn't have any files in its directory - which is definitely an improvement. VS definitely looks pretty slick - you can switch the "compliance-mode" of your page from IE6 to XHTML Transitional. When you switch modes, it changes the code completion attributes - so you'll get much more options for IE6 than you will for XHTML. It also has full code-completion for JavaScript - and the options are based on your mode of operation. Accessibility checking is also included - you'll actually get underlines for elements that don't have all the required attributes. Java needs an IDE like this soooo bad it's not even funny. Can you imagine having full page/HTML validation and code-completion based on doctype?!

Scott did a master/detail screen with VS and made it look damn easy. Most of the audience oohed and aawed. It's cool and all, but the code in the .aspx pages is a lot like JSF - there's hardly any HTML in the pages! It does seem to have much better support for skins and themes - you can easily change the look and feel right in the IDE and there's a whole bunch of built-in themes. The one thing I really like is the page-backing classes have a pre-init() method that can use to initialize properties. JSF really needs something like this. It's a shame that frameworks like Shale have to provide this and it's not a part of the core JSF framework.

ASP 2.0 Beta 2 will ship in the next month. Heh, Scott just gave a demo adding security and authentication in about 20 seconds - complete with Remember Me. There's actually a "Login" control that doesn't everything for you, including validation. The membership system (database) will actually get generated on-the-fly. He follows it up by creating a signup page that creates new accounts in under a minute. I can't really bash on this, writing authentication for Java webapps is definitely harder than it needs to be. Of course, if you use AppFuse, you don't have to write it at all.

Time to run and catch up with my sister - have a good weekend y'all!

Microsoft's Agenda at the Competitive Influentials Summit

They said I could blog everything about the conference I'm going to tomorrow, so let's see how far they're willing to go. A lot of folks have asked me what the agenda is, and until now - I've had no clue. However, today I was sent an e-mail and I'm happy to let y'all know what's going to happen. One thing I noticed is that the Word document's title was "Competitive Influentials Summit". Heh - I guess I'm an "influential" now.

I'm really looking forward to this event. I think we're really going to get wined and dined, and maybe even learn something. It'll be the first time in my life that someone will be picking me up from the airport with a "Raible" sign. I'm leaving at noon on Thursday to do a little St. Patty's day celebrating with my sister (it's her birthday), so I'll miss the "Open Source and Microsoft" session. Hopefully someone else will blog that so we see what they're thinking.

Spring MVC vs. WebWork Smackdown at OSCON

Matthew Porter and I are going to try something a bit different at this year's OSCON. Rather than just getting up in front of the crowd and spewing our technical know-how, we're actually going to make a go at providing some entertainment. I've been to a lot of conferences and I'm tired of just watching someone talk - I'd rather see a good presenter over a knowledgeable presenter. This has inspired our OSCON 2005 talk:

This presentation has a unique delivery style. Rather than one person doing a comparison, there are two presenters - each which is an expert in the framework they're defending. The presentation is delivered as a friendly comparison/debate, which hopes to add some humor in to make it fun for the audience.

Java web developers often have a difficult choice when choosing a web framework these days. There are currently more than 35 open-source Java web frameworks available. How do you which one to use for your project? This presentation picks two of the most popular frameworks, Spring MVC and WebWork and compares and contrasts their features. Topics will include:

1. View options - i.e. Velocity, JSP, HTML Templates (ala XMLC), etc.
2. Testability - How easy is it to unit test with JUnit, with examples
3. Type conversion - i.e. Date, Integer, etc.
4. Validation - How do do it, stengths and weaknesses
5. Tools Support
6. Strengths and Weaknesses

Now the pressure's on - we have to both teach and entertain the audience. Please feel free to post your experiences with Spring or WebWork and why you think one is better than the other.

Is Laszlo a waste of time?

According to Rife founder Geert Bevin, Laszlo ain't all it's cracked up to be:

It's a shame, I really had huge expectations about Laszlo and even tried to sell it to a customer. I'm glad that project was cancelled or I would be in deep trouble.

Under normal circumstances, I'd dismiss this as FUD, but Geert sounds like he did his homework on this one.

Ajax webapps are cool, but non-javascript versions still needed

I think we can all probably learn a lesson from Google. I've heard that GMail is the "gold standard" for Ajax applications. If that's the case, then you should note that they've recently added a "basic HTML" link to the bottom of their pages. With this link, you can view your e-mail using the old way: Yahoo-style, no-JavaScript-needed. My guess is they added it because of demand, or simply to compete with other providers who have this feature. I think it's a good lesson though: use Ajax features in webapps where appropriate, but don't make JavaScript necessary to use your app.

A couple of Ajax features I've been thinking of developing:

• Saving forms with XMLHttpRequest: just display a success message at the top, and switch the "Cancel" button to "Done". Since the form's content doesn't change, this seems like a reasonable use of the technology.
• Switching out entire "content" <div> elements. Most of my apps have a <div id="content">, so it'd probably pretty easy to just replace that in response to button and link clicks. Of course, the hard part is having the requested server-side object load the view template, process it, and send back the content. This is probably more trouble than it's worth.

[TSSS] BOFs, Booze and Benitar

On Friday night, I attended all three of the 7:30 Birds of a Feather sessions. The first one I went to was Spring, where Rod talked about what's coming in Spring 1.3. Rod did a 25 minute presentation on the new stuff and then opened the floor up to Q and A. The session was well attended and I skipped over to the Tapestry/Trails BOF when the Q and A started.

I was surprised to find that very few folks where at the other BOFs. While the Spring BOF had 50+ attendees, the Tapestry one only had around 15-20 and JSF had around 8. I quickly left the Tapestry/Trails BOF when Chris started walking through his Trails Video. He was doing a live version of it, and since I'd already seen it, I figured I wasn't going to learn anything new. I've also been following Trails since it first started, and was more interested in talking about Tapestry.

I walked into the JSF BOF as Ed was talking about JSF 1.2 and what's next for JSF 2.0. This was good timing as I had a few suggestions for 2.0: HTML Templates like Tapestry, bookmarkability (don't make everything a post) and thinking about tools like Tiles and SiteMesh. While neither tools is part of the spec, I think they should be remembered in case there's an opportunity to make integrating with them easier. Ed did mention that JSF 1.2 has pretty much solved the content-interweaving problem, so putting HTML in your JSF JSPs should be better supported.

The very interesting part of this BOF is that Ajax capabilities are very much on the radar for JSF 2.0. They plan on providing native XMLHttpRequest capabilities. My suggestion for this was to provide the setup and registration of requestable class methods as part of the framework, and leave writing the JavaScript to the developer. This was a good BOF and I'm pumped to see that JSF is embracing the next-gen way of developing webapps. Let's just hope JSF 2.0 is released this year and not 2 years from now.

After the BOFs, I joined Matt and Jim to wait for one of Matt's buddies (Scott) to come into town. After he arrived, we headed over to the OpenSymphony open bar at the Bellagio. There, I got to meet Patrick Lightbody and enjoyed several beers and good conversation with the likes of Seth Ladd, Thomas Risberg, Mike, Dion and Christian.

After the open bar closed, Jim, Matt, Scott and I headed just off the strip to the Gold Coast Casino. Matt and Scott wanted to find some poker (tables had a 2-hour wait on the strip) and Jim and I wanted cheap Blackjack. We were pleased to find $5 tables and stayed there for several hours. I don't know what time we headed back to our hotel, but I'm guessing 1 or 2. The rest of the night was pretty funny. Jim and I gambled until 7 in the morning at several blackjack tables. Our hotel had this "celebrity theme", so we had dealers like Pat Benitar and Stevie Wonder throughout the morning. Both of these dealers were great and I got "hooked up" on several occasions. There were at least 10 times where I asked for a card and they didn't give it to me (after which I won b/c they busted). We ended the night at 7:00-7:30 with 5 crisp $100 bills in my pocket. Total cost of the whole trip: $100. Not bad eh?

Getting home yesterday was quite an adventure. After going to bed at 7:30, I woke up by some miracle at 11:00. I don't know if I had a wake up call or what, but my buzz was still in full swing. I caught a cab and headed to the airport. I paid the cabbie with a $25 chip, which he didn't like, but after I told him to keep the change (it was a $6 cab ride) - he happily obliged. At the airport, I took a nap while waiting for my flight to board and almost missed it. They called my name over the intercom b/c I was the only passenger left to board. Luckily, I was awake and made the flight. Upon arriving in Denver, I walked to my car and promptly locked my keys in the trunk. The airport officials got them out for free and I made it home to a very happy family around 6:00 p.m. It's good to be home.

[TSSS] Days 1 and 2

I'm sitting in the EJB3 BOF right now. The room is packed, but it seems most folks are uninterested and the moderators are just talking amongst themselves. Seems like a good time to blog since this BOF doesn't interest me whatsoever. Yesterday, I arrived at 8:00, took a cab to the Imperial Palace (where we're staying) and then headed over to Caesar's for the conference. I registered, assured we could drink beer during the sessions, and attended the (rather dry) keynote. Hani has a good synopsis of this talk.

After the keynote, I went to Patrick and Jason's WebWork talk for about 10 minutes. I soon realized it was an intro to WebWork and left to try and learn something. I went to Craig McClanahan's talk on "The Development of Web Application APIs and Standards for Java." His talk was pretty good, and covered "de facto" versus "de jure" standards. De facto standards are ones that evolve from the community through widespread usage, whereas de jure standards are imposed on the community (like JSF). Again, Hani has the full scoop on this talk.

Next, I went to Dion and Ben's talk on Ajax applications. They talked about XHTML/CSS and how XMLHttpRequest makes rich client-side applications possible. I think the whole Ajax thing is pretty funny. It's something that's been available for several years and my guess is most folks just didn't know about it. I've been using XMLHttpRequest for a couple of years now, and it's interesting to see it become popular all of a sudden. It's quite nice actually. I've been writing HTML/JavaScript for over 10 years, so I find Ajax development pretty easy. I hope to add support for Ajax-type features in AppFuse before this summer.

I wonder when/if the community will realize the power XSL processing in the browser? Since we're all developing XHTML applications now, our pages are XML and we could easily start leveraging client-side XSLT to do some pretty cool stuff. With a client-side XSL sheet, you could do page decoration (like SiteMesh) just by adding one line to your pages. For example:

<?xml-stylesheet type="text/xsl" href="styles/global.xsl"?>

The only problem with client-side XSLT is your pages have to be well-formed XML or everything bombs. With HTML, if you screw something up, chances are the browser will still render it correctly.

After the Ajax review, I had lunch and headed down to the Casino for some beers and gambling. I came back in time for Rod's "Why J2EE Projects Fail." It was a good talk, but there wasn't any revolutionary or new information provided. After his talk, I was motivated to learn more about Web Services Security, but instead opted for beers with Crazy Bob, JIRA Mike and Neon Dion. A couple of beers turned into several, and I found myself having dinner with the SourceBeat guys (Bill, Matt and Jim) a couple hours later. Steak and Lobster was my plate of choice and it tasted quite good. The rest of the night was spent gambling, drinking and harassing Pai Gow Poker dealers. We had breakfast around 2 and made it to bed by 3. Total cost of the trip so far: $300.

I slept in until noon today, after which Jim and I headed back to the conference for some lunch and afternoon sessions. Lunch was good and followed by Oracle demoing JDeveloper and coding EJB3 and JSF with it. I've often wondered about the cost of Oracle's ADF JSF implementation and actually got an answer from one of the attendees. I think he was an Oracle employee but he basically said you have to buy at least 1 copy of JDeveloper ($999) and you get a runtime license for ADF Faces as part of that. That sucks because Oracle's implementation looks like one of the most full-featured ones available. Why should I have to buy a tool I'll never use just to use ADF Faces?

After lunch, we attended Rod's "Advanced Spring Framework" and Craig's "JSF: Dead on Arrival or Raging Success". Rod's Spring presentation covered some advanced Spring features: autowiring, inner-beans, lists, instantiation choices, factory beans and template bean definitions. The presentation was good, and I was pleasantly surprised to find I knew most of the things he covered. Colin spoke about using JDK 5 annotations for transaction demarcation and Keith talked about Spring Web Flow. The Spring Web Flow stuff looks cool, especially since the other framework developers are listening and liking what they see. Craig even mentioned that he'll probably ditch what he's put together in Shale and use Spring Web Flow instead.

Craig's talk about JSF was rather boring, but most of these sessions are (mainly because there isn't a whole lot of new information). Craig did manage to pimp Java Studio Creator a bit. I find JSC demos to be quite funny since it hides so much code with code-folding. In the demo, Craig showed us a 10-line Java class that made JSF (and JSC) look like good stuff. Jim and I noticed code-folding was turned on and the class was actually 120 lines long! This is more of a problem with JSC then JSF IMO. The one nice thing about this talk was learning that a JSF 2.0 BOF is tonight. The main goal of the BOF is to see what the community wants in 2.0. I hope to attend and express a desire for HTML templates like Tapestry has.

Tonight kinda sucks because all the good BOFs (Spring, Tapestry and JSF 2.0) are at the same time (7:30). I'm hoping to hop around between them and get some good networking in. After the BOFs, OpenSymphony is hosting an open bar - so that should be a good time. Hopefully we can scare up a few free carbombs. For more blogs and coverage of the conference, see the TSSS 2005 blogger list.

Simplifying XmlHttpRequest with JSON-RPC

In my day job, we decided to use a little XMLHttpRequest lovin' to populate one drop-down from another. This is my review of JSON-RPC, an open source JavaScript library and servlet for simplifying XMLHttpRequest. I considered integrate Direct Web Remoting (DWR) as well, but its java.net site was down the day I needed it. I started out with JSON-RPC 0.7, which caused some conflicts with Commons Validator client-side validation. This was fixed in the 0.8 release. JSON-RPC takes a little more setup than I care for, but it's pretty easy nonetheless:

1. Download the 0.8 release from http://oss.metaparadigm.com/jsonrpc-dist/json-rpc-java-0.8.tar.gz.
2. Add the JAR to your project and the webapps/jsonrpc/jsonrpc.js to your projects' "scripts" folder. Include this file in your SiteMesh decorator or Tiles layout. If you're not using SiteMesh or Tiles, it's high time you started.
3. JSON-RPC currently requires that you register each class you want call methods on. In our project, I registered a Spring bean (LookupHelper) that's a singleton with references to Maps in the ServletContext. Then we used JavaScript functions to call JSON-PRC and look up units for a plant, and vice versa. I'm not going to put the LookupHelper class here - you'll have to trust its methods return a single String or a comma-separated list of Strings. To register this bean with JSON-RPC, I created an HttpSessionListener and configured it in web.xml.

   /**  * UserListener class used to add/remove session attributes when  * a user first logs in.  Mainly for JavaScript Remote Scripting stuff.  *  * @author Matt Raible  */ public class UserListener implements HttpSessionListener, HttpSessionAttributeListener {      private final Log log = LogFactory.getLog(UserListener.class);     public final static String BRIDGE_KEY = "JSONRPCBridge";     /**      * Initializes LookupHelper singleton with values needed for lookup      *      * @param event the HttpSessionEvent to grab session information from      */     public void sessionCreated(HttpSessionEvent event) {         // Find the JSONRPCBridge for this session or create one         // if it doesn't exist. Note the bridge must be named BRIDGE_KEY         // in the HttpSession for the JSONRPCServlet to find it.         HttpSession session = event.getSession();         JSONRPCBridge jsonBridge = new JSONRPCBridge();         jsonBridge.setDebug(true);         session.setAttribute(BRIDGE_KEY, jsonBridge);     }     /**      * Destroys LookupHelper      *      * @param event the HttpSessionEvent to grab session information from      */     public void sessionDestroyed(HttpSessionEvent event) {         if (event.getSession() != null) {             event.getSession().removeAttribute(BRIDGE_KEY);         }     }     public void attributeAdded(HttpSessionBindingEvent event) {         if (event.getName().equals(BRIDGE_KEY)) {             HttpSession session = event.getSession();             // register LookupHelper so we can call methods on it             ApplicationContext ctx =                     WebApplicationContextUtils                     .getWebApplicationContext(session.getServletContext());             // check for null so we don't have to initialize Spring in tests             if (ctx != null) {                 log.debug("Registering lookupHelper for XmlHttpRequest...");                 JSONRPCBridge jsonBridge =                     (JSONRPCBridge) session.getAttribute(BRIDGE_KEY);                 jsonBridge.registerObject("lookupHelper",                                           ctx.getBean("lookupHelper"));             }         }     }     public void attributeRemoved(HttpSessionBindingEvent event) {         // don't care     }     public void attributeReplaced(HttpSessionBindingEvent event) {         // same as attribute added         attributeAdded(event);     } }

4. After this setup was complete, I was able to add the following JavaScript to the bottom of my JSP. These are functions that our drop-downs call to populate each other, and keep their options in synch.

   <script type="text/javascript"> var jsonurl = "${ctx}/jsonrpc"; var jsonrpc = null; var unitDropDown = document.getElementById("equipmentName"); function filterUnits(plantDropDown) {     var plantName = plantDropDown.options[plantDropDown.selectedIndex].value;     if (plantName == "") {         reloadUnits("");         return;     }     try {         jsonrpc = new JSONRpcClient(jsonurl);     } catch(e) {         alert(e);     }     // Call a Java method on the server     var units = jsonrpc.lookupHelper.getUnitsForPlant(plantName);     setUnits(units); } function reloadUnits(value) {     if (value == "") {         try {             jsonrpc = new JSONRpcClient(jsonurl);         } catch(e) {             alert(e);         }         // Call a Java method on the server         var units = jsonrpc.lookupHelper.getAllUnits();         setUnits(units);     } } function setUnits(units) {     var unitArray = units.split(",");     unitDropDown.options.length = 1;  // keep "All" option     for (i=0; i < unitArray.length; i++) {         unitDropDown.options[unitDropDown.options.length] =             new Option(unitArray[i], unitArray[i]);     } } </script>

The hardest part of using JSON-RPC is setting it up. We only experienced minor issues with Commons Validator, but since the JSON-RPC 0.8 release - everything has worked great, on all browsers we need to support. The only thing I don't like about this library is that you have to register objects for each user's session. I briefly looked at DWR and it looks a little cleaner - especially b/c of its Spring integration. The next time we need XMLHttpRequest, we'll probably use DWR just to compare the two.