Hi Guys, I have a set of questions that I need to get clarified before I can put forth a proposal of implementing Beehive in my company's web based project (If I do propose the usage of Beehive and if there are problems later on, I'm the one who's going to get a lot of flak for it), so please do help.
The application is an N Tier web-application with current proposal of using Struts, Validation framework (with both user defined validations and the ones included in validation-rules.xml), DynaActionFormBeans etc. for front-end and session façade, service locater, helper classes etc. in the middle tier and Stateless session beans and Abstract Factory DAOs for the back end. 1) One of the chief reasons for the scepticism over Beehive is that folks think that Beehive splits the controller (in different controller files) which makes the effort of tracking from one flow to other that more tough. Whereas Struts has a central struts-config.xml file which can be easily used for tracking the flow, there is no such single file in Beehive and this could lead to more time in tracking the flows than in Struts. 2) Beehive does not have a configuration file which could be changed without the need for compilation. Such files go a long way in meeting customer demands in case of a quick business flow change. For example, in Struts a change in scenario A to scenario B [JSP flow] will be so easily changeable by tweaking some tags in struts-config.xml here and there, whereas in Beehive the same activity may lead to multiple sub-activities like tracking the flow in individual Java Controller files, changing the Java code and their subsequent compilation. 3) Can we make DynaActionFormBeans in Beehive or do we have to make explicit Java Beans as form beans each time a flow is decided. 4) Can we use a validation framework with user defined validations. 5) How easy or tough is the task of rendering internationalized message in Beehive. Thanks, Vaibhav
