Why Two Servers?

The perl-based OME Server has been developed and regularly released since 2000. It provides image data and metadata visaualization, management, and analysis facilities. It is actively developed and supported.

Since 2003, the Dundee group has been developing a remote Java client for image visualization and management. The tool, codenamed Shoola, had limited functionality because the remoting interfaces available in the perl-based OME Server could not provide performant data transfer.

Starting in 2004, the Dundee group began exploring an alternative architecture for an OME data server. The project, codenamed OME Remote Objects (OMERO), has built a Java Enterprise-based server application. This server provided a more performant remoting interface, allowing development of more ambitious functionality in a remote application. For performance reasons, we have stopped developing Shoola2.x for the OME Server and have instead been developing remote clients for the OMERO Server.

We recommend that users employ the Web UI for the OME Server and the Shoola-based OME Client and Importer for the OMERO server.

Currently the OME and OMERO servers are incompatible, due to differences in the underlying data model that were implemented in OMERO. Fundamentally, the design requirements for the two servers are different, and the systems solve very different problems. A conversion tool that will read and OME Server instance and write data into an OMERO Server has been built. However, rather than work with converters, this situation gives the OME project the chance to explore an important new type of functionality, the linking of databases. This is important as we move towards our goal of linking other data models and structures--for instance, biological ontologies and RDF. This will allow us to provide a single integrated data management system that leverages the advantages of both systems. Details on this work will be available towards the end of Q1/2007.