Skip to content

v0.7.0-beta

Choose a tag to compare

@mykhailo-kuchma mykhailo-kuchma released this 09 Sep 14:54
· 1330 commits to master since this release

Common

  • Introduced new abstract olp::thread::TaskScheduler to be used for async context; provided default thread pool implementation.
  • Added UserAgent HTTP header to mark all triggered requests.
  • Added China Lookup API URLs.
  • Introduced new HTTP abstraction layer olp::http::Network without singleton usage. Platform dependent implementations adapted and aligned with coding style.
  • olp::network::Network with according implementations and helper classes removed.
  • Changed olp::client::CancellationContext to shared pimpl so we can remove the std::shared_ptr wrapper when using it across multiple requests.
  • Improved logging in core components.

olp-edge-sdk-cpp-authentication

  • Breaking Change - Use new HTTP layer as input parameter for all requests. This is a mandatory parameter and users must provide it.
  • Breaking Change - Use new olp::thread::TaskScheduler; provided as input parameter.
  • Removed usage of raw std::thread and switched to olp::thread::TaskScheduler. If the olp::thread::TaskScheduler instance is not provided, all tasks are executed in the calling thread.

olp-edge-sdk-cpp-dataservice-read

  • Breaking Change - Use new HTTP layer as input parameter for all requests. This is a mandatory parameter and users must provide it.
  • Breaking Change - Use new olp::thread::TaskScheduler; provided as input parameter.
  • Removed usage of raw std::thread and switched to olp::thread::TaskScheduler. If the olp::thread::TaskScheduler instance is not provided, all tasks are executed in the calling thread.
  • Improved logging.

olp-edge-sdk-cpp-dataservice-write

  • Breaking Change - Use new HTTP layer as input parameter for all requests. This is a mandatory parameter and users must provide it.
  • Breaking Change - olp::dataservice::write::StreamLayerClient uses new olp::thread::TaskScheduler as input parameter for async context. If the olp::thread::TaskScheduler instance is not provided, all tasks are executed in the calling thread.