[6.0.0] - 2022-12-07
The latest version of this SDK supports LaunchDarkly's new custom contexts feature. Contexts are an evolution of a previously-existing concept, "users." Contexts let you create targeting rules for feature flags based on a variety of different information, including attributes pertaining to users, organizations, devices, and more. You can even combine contexts to create "multi-contexts."
This feature is only available to members of LaunchDarkly's Early Access Program (EAP). If you're in the EAP, you can use contexts by updating your SDK to the latest version and, if applicable, updating your Relay Proxy. Outdated SDK versions do not support contexts, and will cause unpredictable flag evaluation behavior.
If you are not in the EAP, only use single contexts of kind "user", or continue to use the user type if available. If you try to create contexts, the context will be sent to LaunchDarkly, but any data not related to the user object will be ignored.
For detailed information about this version, please refer to the list below. For information on how to upgrade from the previous version, please read the migration guide.
Added:
- In
com.launchDarkly.sdk
, the typesLDContext
andContextKind
define the new context model. - For all SDK methods that took an
LDUser
parameter, there is now an overload that takes anLDContext
. The SDK still supportsLDUser
for now, butLDContext
is the preferred model andLDUser
may be removed in a future version. - The
TestData
flag builder methods have been extended to support now context-related options, such as matching a key for a specific context type other than "user".
Changed (breaking changes from 6.x):
- It was previously allowable to set a user key to an empty string. In the new context model, the key is not allowed to be empty. Trying to use an empty key will cause evaluations to fail and return the default value.
- There is no longer such a thing as a
secondary
meta-attribute that affects percentage rollouts. If you set an attribute with that name in anLDContext
, it will simply be a custom attribute like any other. - The
anonymous
attribute inLDUser
is now a simple boolean, with no distinction between a false state and a null state. - Types such as
DataStore
, which define the low-level interfaces of LaunchDarkly SDK components and allow implementation of custom components, have been moved out of theinterfaces
subpackage into a newsubsystems
subpackage. Some types have been removed by using generics: for instance, the interfaceDataSourceFactory
has been replaced byComponentConfigurer<DataSource>
. Application code normally does not refer to these types except possibly to hold a value for a configuration property such asLDConfig.Builder.dataStore()
, so this change is likely to only affect configuration-related logic.
Changed (requirements/dependencies/build):
- The SDK no longer has a dependency on SLF4J. It will still use SLF4J as the default logging framework if SLF4J is in the classpath, so it is the application's responsibility to declare its own dependency on SLF4J, as any application that uses SLF4J would normally do.
- Applications that use the database integrations for Redis, DynamoDB, or Consul must update to the latest major versions of the corresponding packages (
launchdarkly-java-server-sdk-redis-store
, etc.).
Changed (behavioral changes):
- If SLF4J is not in the classpath, the SDK now uses
System.err
as its default logging destination. See "requirements/dependencies/build" above. - The SDK can now evaluate segments that have rules referencing other segments.
- Analytics event data now uses a new JSON schema due to differences between the context model and the old user model.
Removed:
- Removed all types, fields, and methods that were deprecated as of the most recent 5.x release.
- Removed the
secondary
meta-attribute inLDUser
andLDUser.Builder
. - The
alias
method no longer exists because alias events are not needed in the new context model. - The
inlineUsersInEvents
option no longer exists because it is not relevant in the new context model.