The goal of the Open Notebook Project is to create cross-platform client-side application to assist the end user of a web-based scientific resource in storing and maintaining returned data. It draws on the assumption that many web-based scientific information resources will increasingly operate a dual-mode service via a conventional web interface and SOAP-based access. An initial list of proposed features includes: Cross platform usability - based on Java or Mozilla A basic infrastructure for local data storage in XML format A basic and efficient search mechanism of the local XML database Remote SOAP-service (Web Services) invocation ability Automated (i.e. unattended) update of returned data from web queries via direct interaction of the application with a web service interface Ability to annotate stored data Automatic correlation of annotations with returned search results Query persistance between sessions Advanced data representation using specific add-in modules. Advanced software modules for the framework might encompass community-based features that could include: A real-time messaging interface and networking mechanism Real-time peer-to-peer querying to find annotations to specific data records between users of the Open Notebook (in effect, a scientific Napster) Scientific communities; real time messaging between different user groups.