Context-Aware Encoding and Delivery in the Web

While standard HTTP caching has been designed for static resources such as files, different conceptual extensions have made it applicable to frequently changing data like database query results or server-generated HTML content. But even though caching is an indispensable means to accelerate content delivery on the web, whether or not cached resources can be used for acceleration has always been a binary decision: a cached response is either valid and can be used or has been invalidated and must be avoided. In this paper, we present an early-stage PhD project on a novel scheme for content encoding and delivery. Our primary goal is minimizing the payload for client requests in the web by enabling partial usage of cached resources. We discuss related work on the topic and analyze why existing approaches have not been established in practice so far, despite significant gains such as reduced bandwidth usage and loading times for end users. We then present open challenges, derive our research question, and present our research goals and agenda.