Releases: TouK/nussknacker
Releases · TouK/nussknacker
Release 1.18.0
Highlights
End-user
- New Activities panel, replacing Versions, Comments and Attachments panels. Now you can browse all scenario activities on one chronological list.
- Added scenario labels. You can now organize your scenarios and find different groups of scenarios more easily.
- SpEL: added navigation through fields inside variables typed as Unknown. You can now access the data inside a variable, even if Nussknacker doesn't know its exact type during scenario authoring.
- SpEL: added conversion methods to cast or convert between data types (e.g.
String
toInteger
). - SpEL: various enhancements, like
#CONV.toJson
and#CONV.toJsonString
methods, new#BASE64
helper, possibility to treat arrays as lists, and more. - Various UX improvements, including new UI for aggregation definitions and better validation handling in ad-hoc tests.
Administrator
- Flink upgrade to 1.19.1. Note: it is possible to use Nussknacker with older versions of Flink, but it requires some extra steps. See Migration guide for details.
- Performance optimisations of the serialisation of events passing through Flink's
DataStream
s.
All changes
For a detailed list of changes see Changelog and Migration guide.
Release 1.17.0
Highlights
End-user
- Flink Catalogs integration. Thanks to Catalogs, Nu can be used to act on data stored in Data Lakehouses (check our tutorial on Apache Iceberg integration).
- Improvements in batch mode, including joins, testing (dry runs), and general usability.
- Live counts. You can now watch counts of events as they flow through a scenario, with user defined update interval.
- Autocompletion now works better in invalid expressions. When working on an expression, you don't have to keep it valid all the time for autocompletion to work.
- SpEL collections usability improvements. Arrays and Lists are now unified (no need for conversions) and we also added more collection manipulation functions.
Administrator
- Performance optimizations around Avro processing that reduce CPU utilization significantly.
- Components configuration reloading without Designer restart. You can now reload the Components configuration on demand, using an endpoint, without restarting the whole application.
All changes
For a detailed list of changes see Changelog and Migration guide.
Release 1.16.3
All changes
For a detailed list of changes see Changelog
Release 1.16.2
All changes
For a detailed list of changes see Changelog
Release 1.16.1
All changes
For a detailed list of changes see Changelog
Release 1.16.0
Highlights
Administrator
- Added Exactly-Once semantics for event processing in streaming mode.
- Added impersonate mechanism. When an external system calls the Nussknacker API with a technical user, the impersonated user can be passed and the performed action will be carried on behalf of the impersonated user.
- Nussknacker Quickstart has been refurbished and is now based on installation example. Using Nussknacker Quickstart is now even clearer and easier.
All changes
For a detailed list of changes see Changelog and Migration guide.
Release 1.15.3
All changes
For a detailed list of changes see Changelog
Release 1.15.1
All changes
For a detailed list of changes see Changelog
Release 1.15.0
Highlights
End-user
- Experimental support for the batch processing mode, based on Flink’s Table API and Catalogs. It allows processing data at rest and opens up new features such as integration with Lakehouses
- New color theme, more consistent, readable, and more violet!
- Added node search in the scenario editor - you can now easily find specific nodes based on their properties
Administrator
- Support for Flink 1.18
- Added support for scenario migration between different Nussknacker versions. From now on, it will be possible to migrate scenarios using the migration API
- New one-command Docker installation script for setting up an example Nussknacker environment on-premise. This example can be a good reference for your own on-premises production setup, as well as for exploration and testing
All changes
For a detailed list of changes see Changelog and Migration guide.
Release 1.14.0
Highlights
End-user
- New Decision table component, that allows expressing multi-condition decisions in a straightforward way
- Scenario processing modes - streaming, request-response, or batch (in the next release!) - made a first-class citizen in Designer, to help users choose how to process the data
- Multiple scenario and fragment validation improvements
Administrator
- Added support for different topic naming conventions in Azure Schema Registries
- Bumped Flink to 1.17
Component developer
- Changes and improvements in Component API for custom component development
All changes
For a detailed list of changes see Changelog and Migration guide.
New Contributors
- @lukasz-bigorajski made their first contribution in #5334
- @Elmacioro made their first contribution in #5327