International Business Machines Corporation

Stephen Hurst

Ian Mark Hodges

Andrew Wright

Payal Mehta

Daniel Jeffrey

Sarah E Jackson

Stephen Evans

Russell P Finn

Michael Vrakopoulos

Peter Griffin

A computer-implemented method includes identifying a bad fix to a software product, where the software product includes a plurality of parts, and where the bad fix presents an error into a first part. A plurality of fixes to the software product, including the bad fix, are representable by a dependency graph indicating dependencies among the plurality of fixes. The dependency graph is recursively traversed to identify a subset of fixes that are dependent on the bad fix, where each fix in the subset of fixes includes a respective set of parts, and to add a reference to a respective most recent version of each part of each fix in the subset of fixes to a parts datastore. A remedial part and each part referenced in the parts datastore are packaged into a remedial fix, where the remedial part includes a correction to the error in the first part.

Remedying bad service fixes having subsequent dependent fixes

A computer-implemented method, system, and computer program product for managing history metadata of a file. The method may include accessing a first file with delta information. The method may also include determining a portion of the first file to be moved. The method may also include reviewing the delta information of the first file for changes to the portion in the delta information. The method may also include, in response to reviewing the delta information, generating a portion delta history for the portion. The method may also include inserting the portion in a position in a second file, where the portion is inserted with the portion delta history. The method may also include generating new delta information for the second file, where the new delta information includes the portion delta history of the inserted portion.

Management of history metadata of a file

A method and system for processing a plurality of unordered software patches for one or more computer readable software programs configured to be executed in a computer system. The unordered software patches are obtained. The unordered software patches include relationship information for defining a relationship the unordered software patches have with a second software patch. For each unordered software patch, the relationship information of the unordered software patch is analyzed to determine the relationship the unordered software patch has with the second software patch. An entry for each unordered software patch is stored in a relationship data store in the data processing system or computer system. The entry comprises dependency data representative of the determined relationship the unordered software patch has with the second software patch. The unordered software patches are ordered based on the dependency data in the relationship data store.

Software patch ordering

A method of operating a device comprising connecting to an external data source via a network, determining that the network connection has been broken, determining that the network connection has been restored, detecting that a user is accessing an application that requires access to the network connection, and prioritizing the accessed application with respect to the required access to the network connection.

Growing community of inventors

Four Marks, United Kingdom

Stephen Hurst