About

For developers, OpenStatServer is WebServices++.
For users, OpenStatServer is Computational Legos.

OpenStatServer aims to provide clean access to computational modules defined in a variety of computational environments (R, SAS, Matlab, etc) via a single well-defined client interface. OpenStatServer will specify standards-based and machine-readable data structure for storing all of the meta information that is necessary to allow client systems to automatically locate, import, and call computational modules. This same meta-information will support software lifecycle management.

Ultimately, OpenStatServer will allow user-interface vendors to provide users with simple tools accessing a large repository of computational modules. Users with appropriate skills will also be enabled to add new computational modules developed in their language of choice.

The OpenStatServer system:

• Turns computational services into web services,
• Efficiently brokers and allocates computational capacity (e.g. large and small Unix hosts, clusters, grid computing, etc) for ad-hoc interactive as well as long-running computational tasks,
• Makes implementation of applications with computational components easier and faster,
• Increases code re-use (and thus best-of-breed development and selection), by removing integration barriers between components implemented within different toolchains,
• Allows generic integration of computational components (eg implemented in R, SAS, FORTRAN etc) as components in external commercial or bespoke UI, visualisation, workflow and otherapplications (eg Pipeline Pilot, TurboWorks?, Spotfire, etc),
• Encourages and supports the growth, distribution and use of a cross-platorm cross-toolchain library of both Open Source and commercial computational services,
• Eases the problems of data format translations across disperate toolchains,

OpenStatServer enables computational programmers to:

• rapidly develop new computational modules,
• rapidly develop and deploy custom web applications,
• use best-of-breed components selected from multiple tool and computation vendors,
• use rapid development methods based on small iterative steps,
• easily use the existing modules in their own work

Call for contributions

I appreciate all your feedback. If you use OpenStatServer on a real world project or if you have an interesting example or demo, please, drop me a line.

Contact info

Gregory R. Warnes Personal page: www.warnes.net Email: Gregory.R.Warnes@Pfizer.com