Abstract: High-performance streaming installation of software applications can be achieved by pre-fetching chunks of application files. Pre-fetching can be directed by a recipe file that can handle a wide variety of application behaviors. If a missing chunk is encountered during execution of the application, a cache miss record can be created and forwarded to a telemetry server. Cache miss records can be aggregated to generate a revised recipe file so that future installations of the application avoid the missing chunk scenario. Using the revised recipe file can cause fewer interruptions, result in less delay and otherwise improve performance.

Abstract: High-performance streaming installation of software applications can be achieved by pre-fetching chunks of application files. Pre-fetching can be directed by a recipe file that can handle a wide variety of application behaviors. If a missing chunk is encountered during execution of the application, a cache miss record can be created and forwarded to a telemetry server. Cache miss records can be aggregated to generate a revised recipe file so that future installations of the application avoid the missing chunk scenario. Using the revised recipe file can cause fewer interruptions, result in less delay and otherwise improve performance.

Abstract: A dynamic and interchangeable set of application behaviors is implemented upon the same underlying software engine. Downloadable data provider behavior descriptors configure the UI generation application dynamically on demand to meet needs that are unknown at ship time, or otherwise cannot be predicted in advance—inputs, formats, contents, and the optimal user interface or experience, all change over time. A data provider supplies the data source locations, data feeds, poll/pull intervals on feeds, parameter definitions, data binding definitions, lists, groups, UI templates, data transformation logic, resources, and UI templates to plug into the base application engine, which transforms the supplied data to create a UI experience tailored to match the appropriate events and available data over time. The base application engine is agnostic to both the data provider and the input data received.

Abstract: A dynamic and interchangeable set of application behaviors is implemented upon the same underlying software engine. Downloadable data provider behavior descriptors configure the UI generation application dynamically on demand to meet needs that are unknown at ship time, or otherwise cannot be predicted in advance—inputs, formats, contents, and the optimal user interface or experience, all change over time. A data provider supplies the data source locations, data feeds, poll/pull intervals on feeds, parameter definitions, data binding definitions, lists, groups, UI templates, data transformation logic, resources, and UI templates to plug into the base application engine, which transforms the supplied data to create a UI experience tailored to match the appropriate events and available data over time. The base application engine is agnostic to both the data provider and the input data received.

Abstract: Multiple data inputs, to a recording device, including out-of-band data, are used in order to dynamically adjust the start and end times for recordings. These data inputs may include published program schedules, electronic program guide data, real-time data feeds, in-band data markers, and real-time services. In advance of and during an event recording, the media recorder monitors real-time data and determines whether the event will start early, start late, end early, or run past the initial scheduled duration. The media recorder may seek out other scheduled recordings for potential conflicts. If no conflicts are found, the media recorder automatically extends or contracts the length of the recording.