Google Inc.

Akamai Technologies, Inc.

Moka5, Inc.

Andrew D Berkheimer

Ben S Hayden

John A Dilley

Prasanna Meda

Charles Alexander Nelson

Constantine P Sapuntzakis

Michael Kerby Chen

William Edward Weihl

John Whaley

Ankur Jain

Philip A Lisiecki

Michael Youssefmir

Paul Lee

Monica Sin-Ling Lam

Eric Chu

Joel M Wein

Vincent Wei-Kang Chen

Erdi Chen

Won-Suk Chun

Ramesh U V Chandra

Alexander Sherman

Kelvin Kam-Suen Yue

Jonathan J Huang

Michael W Brandt

Shivakumar Littoo Rajaraman

Peter Bradshaw

Jonathan Philip Levine

Richard Benjamin Leider

Brian Palmer

Steven Kollars

Systems and techniques are disclosed for predictively selecting media content items and providing the predicted media content items to a cache. A media client may be in communication with a cache and detect the media content items stored on the cache. Based on the detection, a media content user interface may be modified and may contain the cached media content items or links to the cached media content items.

Local client discovery for content via cache

Methods and devices are described for generating a recipe for creating a clip. A clip request may be received at a server, such as a content server. The clip request may be related to a clip including audio and/or video. The server may include one or more processors and a cache, the cache including chunks. Each chunk may store a predetermined positive number of bytes of data. The server may search the cache to determine whether the clip is to be generated from a recipe stored within the cache. The recipe may include one or more tasks. The server may generate the clip in view of the one or more tasks of the recipe. The server may transmit a clip response comprising the clip to the client device.

Apparatus and methods for generating clips using recipes with slice definitions

Methods and devices are described for generating a recipe for creating a clip. A recipe request can be received at a server, such as a recipe server. The recipe request may be related to a clip including audio and/or video. The server can generate a recipe for creating the clip using one or more slices of a data source. The recipe can include one or more tasks that can include a data-source task which can include a specification of the data source and slice definitions for the one or more slices. The data source can include a video and/or audio data source. A slice definition can include a slice value and a slice offset for locating a respective slice within the data source. The recipe can then be transmitted from the server.

Injection of 3D metadata into a requested video is disclosed. An input video and associated metadata are received by a video hosting site. If the metadata indicates that the input video is a 3D video, the 3D video system extracts the accompanying 3D metadata and stores it in an intermediate format, which is independent of the 3D metadata format and video format of the input video. The 3D video system stores the received video in a standard 2D video format in a video database. Responsive to a video streaming request of the input video, the 3D video system determines the actual 3D metadata format based on the video format of the requested video, and injects a modified 3D metadata associated with the input video into the requested video.

Streaming-time 3D metadata injection

A method for facilitating monetization of an embedded video configured for automated playback includes detecting, via an iframe embedded in a user interface of a third party platform, an indication of a user intent to view the video configured for automated playback. The video is hosted by a content sharing platform and embedded in the user interface of the third party platform. The method further includes providing, to a client device, video related data pertaining to the automated playback of the video in the user interface of the third party platform, and upon receiving the video-related data, validating the automated playback of the video in the user interface of the third party platform.

Two-stage playback: verifying user intent to watch an embedded video

Apparatus and devices are described for generating clips by executing one or more tasks contained in a recipe. The recipe can be received at a server, such as a content server, perhaps as part of a clip request. The server can generate a clip by parsing the recipe. Parsing the recipe can include determining whether a given task of the one or more tasks is a data-source task and, in response to determining that the given task is a data-source task, extracting one or more slices of data from a data source specified in the data-source task and adding the one or more slices to the clip. The data source can include a video data source or a combined video and audio data source. Once generated, the clip can be sent from the server.

Methods, systems, and apparatus, including computer program apparatus, implementing techniques for publishing, subscribing to, or playing live appliances. A live appliance includes a current virtual machine image. In publishing, a proxy file of a live appliance file type is provided to the publisher. The type is mapped to a live appliance player; so that when a proxy file is opened, the current virtual machine image is run. The player automatically binds a writeable file system external to the virtual machine image to the image to provide file storage that is accessible from within the virtual machine image and from a host operating system. The player also creates a subscription to the live appliance on the host computer if one does not exist when the proxy file is run. With the subscription, the player runs the then-current virtual machine image whenever the live appliance is run.

Binding user data files

Sharing live appliances

A file transport mechanism according to the invention is responsible for accepting, storing and distributing files, such as configuration or control files, to a large number of field machines. The mechanism is comprised of a set of servers that accept, store and maintain submitted files. The file transport mechanism implements a distributed agreement protocol based on 'vector exchange.' A vector exchange is a knowledge-based algorithm that works by passing around to potential participants a commitment bit vector. A participant that observes a quorum of commit bits in a vector assumes agreement. Servers use vector exchange to achieve consensus on file submissions. Once a server learns of an agreement, it persistently marks (in a local data store) the request as 'agreed.' Once the submission is agreed, the server can stage the new file for download.

Highly scalable, fault tolerant file transport using vector exchange

Highly scalable, fault-tolerant file transport using vector-exchange

A tiered distribution service is provided in a content delivery network (CDN) having a set of surrogate origin (namely, 'edge') servers organized into regions and that provide content delivery on behalf of participating content providers, wherein a given content provider operates an origin server. According to the invention, a cache hierarchy is established in the CDN comprising a given edge server region and either (a) a single parent region, or (b) a subset of the edge server regions. In response to a determination that a given object request cannot be serviced in the given edge region, instead of contacting the origin server, the request is provided to either the single parent region or to a given one of the subset of edge server regions for handling, preferably as a function of metadata associated with the given object request. The given object request is then serviced, if possible, by a given CDN server in either the single parent region or the given subset region. The original request is only forwarded on to the origin server if the request cannot be serviced by an intermediate node.

System for tiered distribution in a content delivery network

Method and system for tiered distribution in a content delivery network

Growing community of inventors

Boston, MA, United States of America

Andrew D Berkheimer