Skip to content

Releases: apple/swift-nio

SwiftNIO 1.7.3

27 Jun 17:39

Choose a tag to compare

Semver Patch

SwiftNIO 1.6.2

27 Jun 17:38

Choose a tag to compare

Semver Patch

SwiftNIO 1.5.2

27 Jun 17:38

Choose a tag to compare

Semver Patch

SwiftNIO 1.4.3

27 Jun 17:38

Choose a tag to compare

Semver Patch

SwiftNIO 1.3.2

27 Jun 17:38

Choose a tag to compare

Semver Patch

SwiftNIO 1.2.2

27 Jun 17:38

Choose a tag to compare

Semver Patch

SwiftNIO 1.1.1

27 Jun 17:37

Choose a tag to compare

Semver Patch

SwiftNIO 1.0.1

27 Jun 17:37

Choose a tag to compare

Semver Patch

SwiftNIO 1.8.0

31 May 14:42
695afc5

Choose a tag to compare

Semver Minor

  • Added new ByteBufferView type, exposing portions of a ByteBuffer as a Collection<UInt8>. Added ByteBuffer.readableBytesView and ByteBuffer.viewBytes(at:length:) to obtain ByteBufferView objects. (#411)
  • Renamed MultiThreadedEventLoopGroup.init(numThreads:) to MultiThreadedEventLoopGroup.init(numberOfThreads:). Deprecated the old name. (#443)
  • Made HTTPRequestDecoder.init(leftOverBytesStrategy:) and enum RemoveAfterUpgradeStrategy public, which allows users creating custom HTTP pipelines to ensure that removing HTTPRequestDecoder after an upgrade was attempted does not cause unexpected bytes delivery. (#438)

Semver Patch

  • Conformed internal _UInt24 and _UInt56 structures to CustomStringConvertible. (#445)
  • Prevented crashes on macOS/iOS when under heavy load and remote peers close connections before we realise they connected. (#453)
  • Miscellaneous code cleanups and testing improvements. (#412, #441, #442, #447, #454)

SwiftNIO 1.7.2

24 May 13:07

Choose a tag to compare

Semver Patch

  • Removed some unnecessary EventLoopPromise allocations. (#437)
  • Fixed issues where removing a HTTPDecoder could lead to the pipeline consuming raw ByteBuffers, rather than decoded HTTP components. (#430)
  • Fixed an issue where datagram writes would be counted incorrectly, potentially leading to precondition failure. (#431)
  • Resolved a number of re-entrancy issues with HTTPDecoder. (#427)