The current project I am working on is deliverying a Portal Framework for a public sector organisation in the UK. Previous to this I worked on a project delivering an B2E Portal for a Pharma company.
On the current project we have a number of debates between some of the techies about the value of Portals and whether we are using the product in the correct manner.
What we are doing to start with is to deliver a single Portal application and the surrounding framework. Without question if this app was all we were delivering then Portal would be the sledgehammer for the nut. This is very much a functional Portal and would possibly have been better built as a simple app for an app server. However the strategy is to supplement this with other Portals Applications which will mix content and functionality such that the Portal as a whole becomes the definitive workspace for the user population. I believe we are doing the right thing by using Portal and the debate we have had tend to get down to the gritty subject of granularity of Portlets, how do you decide when to develop the app as a single Portlet and when as a collection? We have no doubt been impacted by the immaturity of JSR 168, roll on JSR 268, but I believe we have struck a fine balance between too many Portlets with complex relationships and the snigle Portlet which would not allow users to, in the future, develop a dashboard of their ‘favourite’ Portlets.
The dashboard concept is new to the user population and is something we need to start to address as new Portal Apps are developed…I’ll keep you posted on how this goes.