Abstract: For browser cache cleanup, to consider for eviction a data item stored in a cache of a browser application in a device, a probability that the data item will be needed again during a period after the eviction is computed. A type is determined of a network that will be available at the device during the period. A cost is computed of obtaining the data item over a network of the type, from a location of the device during the period. Using the probability and the cost, a weight of the data item is computed. The weight is associated with the data item as a part of associating a set of weights with a set of data items in the cache. The data item is selected for eviction from the cache because the weight is a lowest weight in the set of weights.

Abstract: To identify a source of a memory leak in an application, a pattern of objects is identified in an object hierarchy of a heap dump, the pattern including an indication of the memory leak. The pattern is matched with a metadata of the application. A static entry in the metadata describes a relationship between a component of the application and an object of a class used in the component. A flow entry in the metadata describes a relationship between a pattern of instantiation of a set of objects corresponding to a set of classes and an operation performed using the application. When the pattern matches the flow entry in the flow section of the metadata, a conclusion is drawn that the memory leak is caused in the operation identified in the flow entry. A portion of a code that participates in the operation is selected for modification.

Abstract: For browser cache cleanup, to consider for eviction a data item stored in a cache of a browser application in a device, a probability that the data item will be needed again during a period after the eviction is computed. A type is determined of a network that will be available at the device during the period. A cost is computed of obtaining the data item over a network of the type, from a location of the device during the period. Using the probability and the cost, a weight of the data item is computed. The weight is associated with the data item as a part of associating a set of weights with a set of data items in the cache. The data item is selected for eviction from the cache because the weight is a lowest weight in the set of weights.

Abstract: A system and method for reducing data usage for wireless subscription plans. An email and an attachment are downloaded to a first device of a first recipient. An attachment is identified on the email by an attachment processing service that ensures that sharing from the first device is based on a credit policy. That there is more than one recipient for the email and attachment is determined. That a second device of a second recipient is in the vicinity of the first device is determined. Verification of the second device is established with an email fingerprint exchange handshake based on a fingerprint of the email for the second recipient that is computed from a hash of content of the email, an identifier of the attachment, and an email id of the second recipient. A second network is utilized to copy the attachment from the first device to the second device.

Abstract: For browser cache cleanup, to consider for eviction a data item stored in a cache of a browser application in a device, a probability that the data item will be needed again during a period after the eviction is computed. A type is determined of a network that will be available at the device during the period. A cost is computed of obtaining the data item over a network of the type, from a location of the device during the period. Using the probability and the cost, a weight of the data item is computed. The weight is associated with the data item as a part of associating a set of weights with a set of data items in the cache. The data item is selected for eviction from the cache because the weight is a lowest weight in the set of weights.

Abstract: For browser cache cleanup, to consider for eviction a data item stored in a cache of a browser application in a device, a probability that the data item will be needed again during a period after the eviction is computed. A type is determined of a network that will be available at the device during the period. A cost is computed of obtaining the data item over a network of the type, from a location of the device during the period. Using the probability and the cost, a weight of the data item is computed. The weight is associated with the data item as a part of associating a set of weights with a set of data items in the cache. The data item is selected for eviction from the cache because the weight is a lowest weight in the set of weights.

Abstract: A technique for assigning a user workload to an application server includes receiving a user request to initiate execution of a workload assigned to a user. An application server is selected, from among multiple application servers, to execute the workload based on compatibility between respective current states of the application servers and the workload.

Abstract: To identify a source of a memory leak in an application, a pattern of objects is identified in an object hierarchy of a heap dump, the pattern including an indication of the memory leak. The pattern is matched with a metadata of the application. A static entry in the metadata describes a relationship between a component of the application and an object of a class used in the component. A flow entry in the metadata describes a relationship between a pattern of instantiation of a set of objects corresponding to a set of classes and an operation performed using the application. When the pattern matches the flow entry in the flow section of the metadata, a conclusion is drawn that the memory leak is caused in the operation identified in the flow entry. A portion of a code that participates in the operation is selected for modification.

Abstract: A method for reducing data usage for wireless subscription plans. An email and an attachment are downloaded to a first device of a first recipient. An attachment is identified on the email by an attachment processing service that ensures that sharing from the first device is based on a credit policy. That there is more than one recipient for the email and attachment is determined. That a second device of a second recipient is in the vicinity of the first device is determined. Verification of the second device is established with an email fingerprint exchange handshake based on a fingerprint of the email for the second recipient that is computed from a hash of content of the email, an identifier of the attachment, and an email id of the second recipient. A second network is utilized to copy the attachment from the first device to the second device.

Abstract: A technique for assigning a user workload to an application server includes receiving a user request to initiate execution of a workload assigned to a user. An application server is selected, from among multiple application servers, to execute the workload based on compatibility between respective current states of the application servers and the workload.

Abstract: A technique for assigning a user workload to an application server includes receiving a user request to initiate execution of a workload assigned to a user. An application server is selected, from among multiple application servers, to execute the workload based on compatibility between respective current states of the application servers and the workload.