On Web annotations: promises and pitfalls of current Web infrastructure

Annotations are a broadly useful mechanism that can support a number of useful document management applications (third-party commentary, design rationale, information filtering and semantic labelling of document content, to name just a few). The ubiquity of World Wide Web content motivates the need for Web annotation systems that are lightweight, efficient, non-intrusive (preferably transparent), platform-independent and scalable. Building such a system using open and standard Web infrastructures (as opposed to proprietary ones) facilitates widespread applicability and deployment. In practice, there are a number of ways to do this, all of which instantiate a common abstract architecture based on intermediaries. This paper described our experiences with client- and proxy-server based implementations of an annotation system architecture. The implementations point to missing elements in the current Web infrastructure that make any implementation of annotation systems less than completely satisfactory. This paper discusses these elements of current Web infrastructure, and potential changes to the Web architecture that might make the implementation of annotation systems more complete.