Flutter SDK 3.5.0

Features

• Upgraded Flutter SDK to use iOS SDK 4.21.1 / Android SDK 4.20.2 / JS/TS SDK 4.9.0
• • In the Web integration, the KameleoonClient now starts initializing immediately upon creation. Previously, initialization only began after calling the runWhenReady method.
    initializes only after runWhenReady method.
• Added support for a New/*Returning- visitor breakdown filter in reports.
• Refactored targeting condition logic (remote data requests are now optional) for:
  • Time elapsed since first/last visit
  • New or returning visitors
  • Total number of visits
  • Number of visits today
• Added support for **304 (Not Modified)*- responses from the SDK config service to avoid redundant updates and reduce traffic when the configuration hasn't changed.
• Added support for using a Custom Data value for traffic bucketing instead of the default visitor code. Learn More.
• Introducing a fallback configuration mechanism via the defaultDataFile parameter. If no cached configuration is available, the SDK immediately uses the provided default. For version control, the SDK prioritizes the default configuration if its dateModified timestamp is more recent than the cached version.
• Added the evaluateAudiences method. This method iterates over all Audiences Explorer segments, evaluates each one, and tracks the segments for which the visitor is targeted using the TARGETINGSEGMENT event.
• Introduced the new getDataFile method. This method returns the current SDK configuration (also known as the data file) used for evaluation and targeting. It is *not- intended for production use to fetch variations for every feature flag in the returned list, as it is not optimized for performance. For that purpose, use getVariations instead. getDataFile is mainly useful for debugging or QA, for example to let internal users manually select a variant for a specific feature flag in production.
• Added a new property **name*- to Variation, representing the name of the variation.
• Added an overwrite flag to CustomData, used as the overwrite parameter during tracking.
• CustomData can now be created using a name, in addition to the existing method of using an index.

Bug fixes

• Changed the order in which *conversion- and *experiment- events are sent. This may lead to more accurate visit-level experiment reporting.