WebDAV

Abstract: Access to WebDAV (Distributed Authoring and Versioning) servers is provided in a manner that is essentially transparent to applications. A WebDAV redirector and related components support file system I/O requests and network requests directed to WebDAV servers identified by URI (Universal Resource Identifier) names, or by a drive may be mapped to a WebDAV share. An application's create or open I/O requests directed to a WebDAV server are detected, and result in a local copy of the file being downloaded and cached for local access. When closed, the local file is uploaded to the WebDAV server. Network-related requests such as for browsing that are directed to a WebDAV server are also handled transparently. WebDAV files may be locally encrypted and decrypted at the file system level, transparent to applications and the WebDAV server, via an encrypting file system that performs local encryption and decryption at the local file system level.

Abstract: Collaborative authoring tools generate network effects, where each tool's value depends not just on the tool itself, but on the number of other people who also have compatible tools. We hypothesize that the best way to generate network effects and to add collaborative authoring capability to existing tools is to focus on the network protocol. This paper explores a protocol-centric approach to collaborative authoring by examining the requirements and functionality of the WebDAV protocol. Key features of the protocol are non-connection-oriented concurrency control, providing an upward migration path for existing non-collaborative applications, support for remote manipulation of the namespace of documents, and simultaneous satisfaction of a wide range of functional requirements.

Abstract: Techniques are presented for allowing clients and servers in a computer network executing the WebDAV protocol to identify a specific version of a specific resource a specific version using a resource tag. The resource can be identified even though it has been changed at a server or at an off line local cache of a client that is disconnected from the network and then later re connected to the network for uploading. Also, a resource UID is presented that will not change despite changes to the URL or the resource tag of the resource. Each resource UID of each resource can be cached locally at a client and can be stored at network server in an index. The index allows the resource to be identified uniquely across a collection in a database at a server, across a database at the server, across the entire server, or across all servers in the network.

Abstract: PeCMan, the Personal Content Manager, is a content access gateway, aggregator, indexing, and sharing tool for all of a user's personally relevant content on the Internet, e.g., all the content a user has created, content that was explicitly shared by others, or publicly available Internet content of specific interest. PeCMan aggregates access to multiple network-based storage systems, services, or other “objects” by their reference. Examples of such objects are photographs, music, e-mail, text documents, or any addressable object on the Internet. The aggregated storage acts as a virtual read-write drive in which PeCMan addresses individual objects by reference, typically encoded in uniform resource locators (URLs). To address content in Web applications such as Google Docs, Flickr, and YouTube, the URL is based on the corresponding application programming interface (API), while content on a user's home machine's disk is addressed via a locally installed software driver supporting Web-based distributed authoring and versioning (WebDAV). This paper describes the motivation, design, and implementation prototype of PeCMan. e 2010 Alcatel-Lucent.