Guus Bosman

Two weeks ago I was starting up a new project, and I wasn't sure if I should introduce the Spring framework. I decided not to. First, the immediate benefits are not clear enough, unlike Hibernate for example, but most of all I didn't want to introduce too many new technologies at once to the development team.

We are using Struts as MVC framework, which is more and more becoming a company standard for web applications, but not everybody on the team has worked with it before. Another new tool for some of the team members is Hibernate. The (existing) database uses stored procedures. It turns out that there are only a few of those and that they're used only for a few specific data insert statements -- not to manage all possible data manipulations. This means there was no data manipulation layer yet and so using Hibernate for the management tool is not a problem.

A last nice feature we're using is a ServletFilter that validates the HTML that we're generating. Not really rocket-science, but a strong and certain way to make sure that the HTML we're creating is standard compliant and valid.

Combined Hibernate and Struts make for a nice and very productive application framework. One of the things I would like to do next time is to look into something to deal with the boiler-plate code that's sometimes necessary for Hibernate. A colleague is looking into Hibernate Synchronizer, but my initial feeling is that a more well known tool such as XDoclet might be a better investment. It's a bit too risky for my taste to build an application on a tool created by single person. His website has been down for a couple of days too -- not a good sign, either.