2009-07-03T17:08:01Z 3349 Home 6 15871 = OSLash : OS Absraction Layer Wiki = OSLash is a cross-platform framework for developing powerful applications with high-performance requirements, by providing utilities for managing common aspects in computing applications like object management, serialization, and concurrency. == Features == * Cross platform Multi-threading API which synchronization tools like mutexes, conditional variables, barriers, count-down latches and intrinsic atomic operations. * Circular buffers for multi-task synchronization, solving the problem of '''1 Producer to 1 Consumer''' and '''1Producer to Multiple Consumers'''. * A powerful '''ThreadPool''' implementation optimized for the '''Parallel-For''' pattern for executing a set of concurrent tasks in parallel. * A highly efficient implementation of common data structures, containers and algorithms like vectors, lists, hash-table sets. Lightweight, portable and also has a very competitive performance performance comparing to c++ STL library. * Memory management for multi-threaded applications with a custom version of the Scalable Allocator as the one implemented in the Intel Threading Building Blocks library. Different implementation, same principle. More readable and maintainable in code. * Management of Global variables with dynamic memory allocation. * Text and string utilities with regular expressions. * Serialization of objects with JSON syntax and XML support. * Time classes and functions for profiling. * Filesystem utilities. * Classes for object management with reference counting and smart pointers. * A robust Math and numerics library for consistent and deterministic calculations across different platforms. Manages number serialization, and also fixed point arithmetics support. <h1><a name='OSLash_:_OS_Absraction_Layer_Wiki'></a> OSLash : OS Absraction Layer Wiki </h1> <p> OSLash is a cross-platform framework for developing powerful applications with high-performance requirements, by providing utilities for managing common aspects in computing applications like object management, serialization, and concurrency. </p><h2><a name='Features'></a> Features </h2> <ul><li> Cross platform Multi-threading API which synchronization tools like mutexes, conditional variables, barriers, count-down latches and intrinsic atomic operations. </li><li> Circular buffers for multi-task synchronization, solving the problem of <b>1 Producer to 1 Consumer</b> and <b>1Producer to Multiple Consumers</b>. </li><li> A powerful <b>ThreadPool</b> implementation optimized for the <b>Parallel-For</b> pattern for executing a set of concurrent tasks in parallel. </li><li> A highly efficient implementation of common data structures, containers and algorithms like vectors, lists, hash-table sets. Lightweight, portable and also has a very competitive performance performance comparing to c++ STL library. </li><li> Memory management for multi-threaded applications with a custom version of the Scalable Allocator as the one implemented in the Intel Threading Building Blocks library. Different implementation, same principle. More readable and maintainable in code. </li><li> Management of Global variables with dynamic memory allocation. </li><li> Text and string utilities with regular expressions. </li><li> Serialization of objects with JSON syntax and XML support. </li><li> Time classes and functions for profiling. </li><li> Filesystem utilities. </li><li> Classes for object management with reference counting and smart pointers. </li><li> A robust Math and numerics library for consistent and deterministic calculations across different platforms. Manages number serialization, and also fixed point arithmetics support. </li></ul><p><br /> </p> 2009-07-07T20:34:15Z 18126