We are contemplating about the technology stack for a rich internet application for a data intensive application. The typical characteristic of the application is to load a big amount of data (say around 40000 records and each record having 8-10 fields) and than work on it on the front end. The user might change the data in any way and than on clicking on save we have to take it back in one go.
Also the data is not plain records but a nested hierarchy of objects (around 6 deep).
At present we are working with GWT and Spring combination with Gilead to handle the cloning mechanism but not seeing good performance.
Some of the options that we are thinking of are: - JNLP - HTML5 feature of data storage in browser
Any insight or past experience would be helpful