v1.5.0

[1.5.0] - 2022-04-14

Added:

• Each release now includes binaries for the arm64 architecture (these were added manually to the 1.4.0 release, but they will now be added automatically).
• Added a test for disabling events.
• Added opt-in "service endpoints" capability for SDKs that support this mechanism.

Changed:

• The tool is now built with Go 1.18.

v1.4.0

[1.4.0] - 2022-04-12

Added:

• Tests for basic HTTP behavior of analytics event posts (request path, headers, etc.).

v1.3.0

[1.3.0] - 2022-03-08

Added:

• Tests for new SDK application metadata properties, enabled by the "tags" capability.

Changed:

• Improved test coverage for private attributes in events.

v1.2.0

[1.2.0] - 2022-02-09

Added:

• Command line options -record-failures and -skip-from.

v1.1.6

[1.1.6] - 2022-02-08

Fixed:

• Analytics event tests no longer care about the order in which events appear in a payload; the order isn't of any significance to LaunchDarkly.

v1.1.5

[1.1.5] - 2022-02-04

Fixed:

• Fixed a bug that could cause the program to crash with a panic when certain tests failed.

v1.1.4

[1.1.4] - 2022-02-03

Fixed:

• Stacktraces now appear consistently for all failures. Previously they only appeared sometimes in console output (in a somewhat different format) and never appeared in JUnit output.
• Debug logging for a subtest now includes log output from components that were created in a parent test.
• In evaluate requests to the test service, valueType is always set.
• Duplicate event posts are ignored by default if they have the same X-LaunchDarkly-Payload-Id header value.

v1.1.3

[1.1.3] - 2022-02-01

Fixed:

• Fixed excessive usage of sockets/file handles due to not always using Keep-Alive for HTTP requests.

v1.1.2

[1.1.2] - 2022-01-28

Fixed:

• Many event-related tests have been rewritten for better separation of concerns, so that if the SDK behaves wrongly in a particular area such as the computation of user properties, the error will be more clearly visible in tests for that area and will not break other tests. Failure messages should now be clearer in general as well, due to changes in how the assertions are done.
• Fixed a bug that prevented the tool from running on Windows.

v1.1.1

[1.1.1] - 2022-01-27

Fixed:

• The "all flags" tests now include test cases for experimentation behavior. There is a known issue in some of the SDKs where the "all flags" data has incorrect properties in these cases, so if contract test jobs start to fail on the evaluation/all flags/experimentation test when using this version, it is likely an actual SDK bug.