Why GWT is a great concept.

One of the most powerful aspects of tools, samples, templates, etc is the fact that most developers learn by example.  So while the last post I put here (Is GWT the future of web development?) went viral on the internet and you heard many rebuttals, affirmations, and other comments, there is one fact that remains – GWT is a very cool technology that “hides” a lot of the complexities of creating state-full web applications that implement Web 2.0 functionality.  It also does another thing, it gives Java developers a bridge over to JavaScript/HTML programming.  By writing Java (which the developer is very familiar with) and outputting a Web 2.0 application it immediately gives the Java developer a one-to-one relationship – or at least GWT’s interpretation.

I think of compilers and tools like this as a point in time technology.  It will be great in the beginning but in the end the developer will rarely “compile” the web application and possibly just edit the output directly – or rewrite it from scratch.  The problem I see is the steep learning curve so the latter part of that  statement may be a while.

I relate things like this to MFC, MFC was great for C/C++ developers but once you learned what MFC did under the covers for the most part many just went around MFC and removed the bloat.

So if Google and the open source community do this right, GWT may stay around for quite a while.  If it ends up creating “bloaded” applications then I think in the end it will not survive.

Lastly, it looks like this approach is gaining popularity.  From the comments there were a couple of open source projects and even a commercial project Vaadin that essentially do the same thing.