Hi, Over the last few weeks we've made some small changes to Maverick 1.0. I'd thought I'd enumerate them here, hopefully either they are already superceeded by Maverick 2.0 or if the changes we made are considered useful they can be rolled into a future Maverick 2.0 relelase? Anyway, the changes are :-
* We couldn't use maverick within a .war file since it makes some calls to getRealPath() on the webapp context. As I understand it if a web app is run directly from a .war file then getRealPath() can return null and getResource() or getResourceAsStream() should be used instead. This required small changes to ConfigLoader.java and PipelineTransforming.java. Grepping through the 2.0 source code it appears that there are still calls to getRealPath(). * We were finding that a lot of our commands would use the same basic controller class but with slightly different options (e.g. retrievals from a database.) Initially, this led to a proliferation of classes which typically just passed some parameters to the base class constructor. To help prevent the number of classes we modified the controller creation so that within the maverick.xml <param name=" " value=" "/> could be contained within the <controller> element. These parameters were then used to "bean populate" the controller in exactly the same way as if they had been passed as URL query parameters. I suspect that the factory mechanism in v2.0 would provide another solution to this problem? * Just before an XSL transformation is used, we check the modification time of the underlying file. If the file has been modified since the template was created the in-memory template is recomputed. This means we can simply change the XSLT file and maverick will automagically detect the changes. If it would be useful we can send patches against the 1.0 release for the above changes. Cheers Guy _______________________________________________ Mav-user mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mav-user
