its to big for use in simple cases like a blog. making a blog in wicket is just hammering nails. making something similar in brix means adding alot more stuff and not needed wrapping. It is per example no need in my blog for changing or creating the templates when the application is running. templates can be predefined and switched in an administration panel.

My final conclusion is that i have not yet seen the best of Brix and that I still need some time playing with it and learning its internals.

I will absolutely recommend Brix for more advanced sites that need an active dialog with designers and webdevelopers. they wont even know that the page is written in Java because all they see is Xhtml!

so what am i suggesting? develop a universal browser framework that takes care of rendering and let people make good looking wrappings for it! Generally it would mean no difference for the end user and party time for web designers (because they would no longer be needed most of the time). and the Java web developers ( count me in) would be a bunch more productive and would generally live a happier life

so i tried to install fedora (that i would classify as developer focused), destroyed my ubuntu partition along the way because i just had to operate the gui while the disk was formatted and then fedora froze. then rebooted and grub didn’t work.

i considered that since i had been smart enough to commit before i began the process (duh) i would just wipe the partition and use it as a data partition.

Next. fedora is installed. first things first. i googled for a reasonable way of installing suns jdk. one hit told me to make direct symbolic links. Even i saw that was not an option. a better way was to use alternatives. i even got Firefox up and running with Java in one line of code. and that is very important. i never managed that in ubuntu. but then again you could say that i installed fedora only to learn how to update Firefox’s libjavaplugin.so with alternatives. which of course is insane but i didn’t like ubuntu anyhow.

I then downloaded eclipse. but as a total silly i chose the 64 bit version. my fedora was x86 so it couldnt execute.

then it was time to go home for the weekend. left with the hope that somewhere on the Internet there is a Linux bundled with the latest sun Java development kit and eclipse enterprise edition. it must be because this is killing me. over and over again.

This is where apache wicket kicks in. It solves my problem to the best. It gives me: complete control of the application logic, code free templates, good abstraction on top of the servlet api and a hell of alot fun! Making an ajaxified submit form? its done.

But is It really the best? Aren’t there strong people behind all of the most widely used web frameworks? havent they been there done that? I expect so.

So, where will I go next? after delving into Spring mvc/webflow, Freemarker and GWT, a new framework? Like WebWork? Or is it a more sensible approach to have an open mind and choose/learn whatever seems to be the most fitting for the current project? yes. of course. fanatics are not doing something they like. they do what they have learned and neglects the fact that there are always better solutions around. Always.

Code is dynamically changing.

function SubmitButton_InitOnClick( id, className )
{
var sb = document.getElementsByName(id)[0];
sb.aspnet_onclick = sb.onclick;
sb.button_className = className;
sb.onclick = SubmitButton_Click;
}

Even if it is a good solution it does not work well with updatepanels because the buttons are rebinded. Thats a topic i have not yet solved. Please tell me if you have a solution to this problem.

I think my wicket CMS will get a move on now please tell me if you want to contribute by sending me a mail to post (at) jarl-andre.com. I am not yet on sourceforge with the project. Thats another project

And then Wicket comes into the picture. It doesnt make the case any better. Actually, wicket is the rant for this article.

Wicket uses abstraction and inheritance to develop interfaces. Mostly like we do with Swing. But does it really make the cut?

With developing web applications as a daily job, boilerplate is always an issue. And with wicket that boilerplate issue increases. Yes i can make reusable components, but that doesnt advocate its use as an out of the box framework. Its very difficult to make a dropdown menu per example whereas in web flow this is just a template.

To conclude: wicket cant stand up to Spring Web Flow when it comes to speed of development. I can make a flow app from Pojo in about one hour that actually works. With wicket i have to code, compile, package and test. In flow i can just edit a flow definition or two.

I Think Ill keep on using Spring for my development.

I couldnt really get the grasp on Web Flow at that point. It was only recently that i discovered how simple and easy this framework really is.

Web Flow is all about control. You will see all the nifty details and can make sure that the page you are managing has the required behaviour. And all defined in an XML file. And another thing; if i ever decide to use another framework my POJO service and domain classes wont be difficult to make use of.

Some advocate that Web Flow has poor binding support. That binding custom classes is not easy/possible. But I say that its a small matter compared to what I gain. Almost everything can be bound with strings and arrays.

I am creating a cms that is taking use of Web Flow, and its a breeze. No more extended controllers!

I really like my current language, but at some times in my life it just sucks. I like web development, and Java doesn’t quite convince me to be a all year round language. For hidden background business logic and patterns where security is a big issue, yeah .. Java is a good starting point. But I really like scripted and parsed languages more, like PHP.

Ovet the last years I have made countless tries on making a blog or home page in PHP, XHTML and CSS. It has mostly always shown me that its really best to use already created solutions, because I can use my skills on other things instead of hacking with one specific thing for a long period of time.

When it comes to Java I must say that it has been really fun to develop all these chat servers and applications that use sockets, web services and the like. But as with most things, they become outdated and you notice that the frameworks does not actually manage to do it quite right. As an example java.io is outdated by java.nio. But then java.nio is so difficult to use. Why? I don’t know and should really like to see a framework for java.nio that kind of resembles the simplicity of java.io. If I don’t find it I will make it myself, and license it with GPL. But its all about time, and I get less and less time as my days goes by.

I love to program! Almost as much as I love my wife!