The podcast, split into two 1/2 hour parts, captures a team of Manning authors and guests discussing (among other things) various Java and non-Java web frameworks. Manning's Marjan Bace assembled the panel, and moderates. The authors and guests include Yehuda Katz, Chad Davis, Max Carlson, Robert Hansen, Bear Bibeault, Chris Richardson, Dan Allen, Emanuel Bernard, Peter Armstrong and Obie Fernandez. Technologies discussed include Seam, Flex, Open Lazlo, Merb, Ruby on Rails, and Hibernate. The debate gets quite lively in part two, and the language is a bit strong in parts. However, it does capture the quandry of a web developer when they ask "What web framework should I choose..." You can subscribe to the podcast feed here: Via RSS: http://chariotsolutions.techcast.com/rss You may also sign up via our podcasts page (contains the iTunes link and a shownotes page): http://www.chariotsolutions.com/downloads/podcasts Direct links to each podcast episode: Part 1 Part 2 Enjoy, Ken Rimple

I am very disapointed with the level of the technicality in the arguments made. How can all these book writers and lead developers be such poor debaters. I learned nothing! I excpeted to hear serious arguments supporting Seam or Struts, but instead I heard nothing. The Ruby guys owned almost everyone else, no one really had any serious counter arguments to their claims. And they didnt make such brilliant arguments either! The Ruby arguments were, Ruby is far way ahead of the competition, the smartest developers work on Ruby and use textmate and macs (those are the real words they used) The counter arguments was, yea, Ruby people are the smartest (!!!) but twitter is slow(!!!) I mean, okay, I am not an expert on so many things, but does this mean that active records is better than hibernate and that deploying a RoR application on a webserver is like more advanced that using ... say glassfish to deploy your web app writen in Struts2 and Grails, are application servers useful at all !!!! Is RoR really the most polished piece of code ever (IN HUMAN HISTORY) Come on, this debate was poor, very poor

Agreed. The arguments were completely juvenile. There was lots of ranting and bashing and emotion, and very little calm and intelligent comparison. The world *needs* a good web framework comparison. There is a maze of choices, lots of misinformation, and these frameworks take huge quantities of time for developers to fully grasp. A quality overview and comparison would be excellent. But this is awful. For example, the Rails representative predictably says: There is no way that anyone in their right mind would do a web project in Java. The only reason to use Java is for legacy reasons... even then it's terrible... That's a valid perspective. He should back it with specific reasons rather than rant how terrible Java is at a vague level and cite "expert" testimonial. I've used Rails. I liked it. However, I am much happier and more productive using Stripes in Java. Why? Honestly, I'm wise enough to admit I don't know Rails enough to make a thorough, intelligent comparison. I just built a few small apps and that was enough to say, I get it, it's cool, but I'm better sticking with Stripes. ASP.NET, JSF, Struts, Stripes are frameworks I've used extensively and can make insightful comparisons between. I've used most of those for dozens of high-pressure projects over 3+ years, I've used them in teams, I've seen other developers fall into common pitfalls, etc. A good shootout would have people who have that kind of deep experience with the frameworks they are comparing and would be able to make useful level-headed evaluations among them.

When I read

The Ruby arguments were, Ruby is far way ahead of the competition, the smartest developers work on Ruby and use textmate and macs (those are the real words they used)

I laughed my ass off. I’ve actually seen these guys come into our JUG with macs and use textmate to generate a simplistic shopping cart type app. When you ask them about interactive debugging, transaction handling, ETL, job scheduling, security, integration, messaging or other features they look back with a blank stare. Enterprise development is much more than generating a front ends for web apps. I’d say maybe 70% of our code has no visual component at all.

YESSS... that's what has always been my sort of question, ok ROR let's you develop simple applications faster... (even though I don't know a whole lot about ROR), but Are you (ROR supporters) telling me that ROR can all the tools to handle transaction mgt, messaging layer, web services consuming/producing, etc.... the list goes on... I would say ROR developers must be those who are developing standalone type of web apps...(another php which is a bit better than php) not doing any enterprise development.

...but Are you (ROR supporters) telling me that ROR can all the tools to handle transaction mgt, messaging layer, web services consuming/producing, etc....

As someone on both sides of the coin (Java systems integration by day, Ruby on Rails by night), you'd be surprised at what is available. Ruby has a very powerful utility for traditional SOAP based web services consuming and production called "Soap4R". This library makes connecting to your Java service provider a snap. I've used that library for writing some test and integration utilities. Axis leaves much to be desired in comparison. ROR is designed around the RESTful URL API out of the box. That is the preferred method for doing web services with Rails, even though it supports automatic WSDL generation for any controller. RoR does support transactions: http://api.rubyonrails.org/classes/ActiveRecord/Transactions/ClassMethods.html . After all, there are enterprise apps written in rails as well. They even have a very nice security guide: http://guides.rubyonrails.org/security.html . I continually find myself frustrated trying to find a similar elegance to these common problems in the Java world and have yet to find them. I've had people really excited about Wicket get me to try it. After the pilot I wrote, I'm not convinced that this is the way web applications ought to be written. It's a great framework for what it does, but it doesn't necessarily fit the way I think about web development. I've attempted to pick up struts on more than one occasion and found it to be offensive. The best framework I've used in Java is one that I threw together that took the basic principles of the original rails routing feature, but implemented in a Java way. There was a package for your controller classes, actions were mapped to methods. After the controller logic was done, the framework delegated to the underlying rendering system (by default JSPs and HTML). It was simple, and didn't get in my way. I'm sure there are a dozen one-offs like it, but it required very little training to understand--which is the entire strength of the rails platform. All I did was leverage its strengths.

Thanks guys, I have not listened to this but I always like to read your comments,,,,

Instead, I think it may be better if each of them talk about the flaws in the framework they're familiar with. This will be more informative for all of us.

Instead, I think it may be better if each of them talk about the flaws in the framework they're familiar with. This will be more informative for all of us.

To that I will heartily agree.

My mistake--the feed url is: http://techcast.chariotsolutions.com/rss

This was the not-Struts, not JSF - well then what discussion. The best take-away from the discussion came from the POJO-in-Action guy Chris Richardson who was concerned with frameworks that tightly coupled or drove domain model design - primarily Seam and Ruby. The domain model will last longer than the various frameworks that come and go. Flex is best if you can dictate Flash player version. (Hard for corporate intra/extranets). Open Laszlo solves the problem by generating code across Flash players and to Javascript as well. Seam like Ruby increases developer productivity without the attitude. GWT looks good since it supports a service layer but has undeveloped/rough components. Now that ext and myGwt have changed licenses and GWT-Ext has lost steam - who knows. This discussion was low value compared to the the other presenetations. Overall a good conference. Other take-away. Don't hire Ruby consultants since the solution to all problems is Ruby.

Andrew, In a way, this sort of presents the Java web framework landscape as it is today--a bit of a mess, with lots of options and no clear front-runner (everyone may argue with me as I duck now)... :) If you want to review the other presentations from ETE, as well as other conferences we've presented on in the past, we have made them all available for free on our website: http://chariotsolutions.com/downloads/presentations