Upgrading to Newer Versions

This page gives instructions to upgrading to newer versions of the Python Tracker

Upgrading to v1.0.0

There are a number of breaking changes and new features in the V1 release, in order to make use of the latest tracker make sure to:

• Tracker namespace is now a required property in the tracker and should be parsed on initialization of a new tracker.
• The track_xxx() functions have been deprecated in favor of the Event class, further information on tracking events using the new API can be found on the Event Tracking page.
• The track() function in the tracker now returns an event id, instead of a tracker instance. Chaining methods using the track() function will no longer work, and methods should be applied to the tracker itself.
• The track() function no longer accepts a Payload object as an argument and instead accepts an Event object.
• The parameters in the Tracker constructor have changed order, to avoid errors we suggest explicitly defining the variables.
• Event subjects and tracker subjects are now combined when an event is tracked on a given tracker. The old API will continue to work, however the contents of the payload in the event will contain additional tracker subject data.
• The Redis and Celery emitters have now been removed from the Python tracker. An example of a redis implementation can be found in the examples section of the tracker repo.

Upgrading to v0.13.0+

There are a number of breaking changes in this release, in order to make use of the new features please make sure to:

• Use the batch_size argument instead of buffer_size in the Emitter initialization.
• Users should no longer use the on_failure method to determine retry behaviors. Retry failed events is now a built in feature which can be configured in the Emitter directly or via EmitterConfiguration object.