Hello Skyline Team | Dojo Community,
DataMiner's C#-based automation, protocols, QAction framework and low-code capabilities have served the platform well over the years. However, Python has become the de facto language for automation, AI/ML integration, data engineering, orchestration and operational tooling across the wider industry.
What is Skyline's roadmap for introducing native Python support within DataMiner, either alongside or as an alternative to C# for Automation Scripts, Protocol Development, GQI extensions, connectors, low-code applications, and operational workflows?
Specifically:
- Is native Python execution currently being considered for future DataMiner releases?
- Would Skyline envisage Python as:
-
- A first-class scripting language within DataMiner?
- A supported runtime for Automation Scripts?
- A framework for Protocol/Connector development?
- An integration layer for AI and Data Science workloads?
- What are the primary technical challenges preventing broader Python adoption today? Security, performance, sandboxing, multi-agent execution, or backward compatibility challenges?
- Are there any plans to expose DataMiner APIs, GQI, DOM, Automation or Data Aggregator components through a Python SDK comparable to today's C# development model?
Hi Samson,
Thanks for the detailed question. Here's where we stand.
Where Python is supported today: connectors You can already use Python inside DataMiner connectors. A connector can include Python connector scripts that collect data and push it into DataMiner, with each connector script running in its own Python virtual environment, which keeps the dependencies of different connectors separate. You can run these scripts on a DataMiner Agent (comes with DaaS out of the box) or closer to the data source using DataMiner Edge. Developers can add Python scripts to a connector solution in Visual Studio (or VS Code) and then package and publish the connector the usual way.
Where Python is not planned right now We have no current plans to add native Python support elsewhere in DataMiner. That includes Automation Scripts, QActions and the protocol framework itself, GQI extensions, low-code apps, and a Python SDK for DataMiner APIs, DOM or Data Aggregator. For those, C# and our existing development model remain the supported route. Python in DataMiner is about connectors, not a general replacement for C#. We're working on letting DataMiner Assistant generate Python scripts that push data into an element, so getting data from a new source becomes even faster.
A new DataMiner API with full Protocol Buffers support We're building a new DataMiner API with full Protocol Buffers (protobuf) support. We expect it to become available in 2027. Because the whole API is defined in protobuf, you'll be able to generate a client in any language with protobuf support, including Python. That doesn't mean we'll provide a Python SDK. But it will make integrating with DataMiner much easier, and Python clients will be able to talk to DataMiner directly.
The protobuf definitions also give AI agents a clear, machine-readable description of the API. Coding assistants can generate reliable client code from them. AI agents can call DataMiner directly to read data, trigger actions and fit into larger automated workflows. .
AI makes C# much easier to get started with AI is also making the choice of language matter less. AI assistants can now generate much of the C# code for you, so you don't need deep C# experience to get going. Our public agent marketplace has plugins for connector and Automation Script development that help you set up a working project quickly. You describe what you need, and the agent produces C# code that follows our development model, which you can then review and refine.
To your specific points
- Native Python execution: Yes, but only for scripts inside connectors. We're not considering it for other parts of the platform at the moment.
- Role of Python: It's a supported way to build the data acquisition logic in a connector. It's not a first-class scripting language across DataMiner, not a runtime for Automation Scripts, and not a replacement for the protocol/QAction framework.
- Challenges: Isolation, managing dependencies, security (for example, handling credentials) and deployment across multiple DMAs are real concerns. We addressed them for connector scripts. Doing the same across Automation, GQI and the rest of the platform is also big change, and it isn't on the roadmap today.
- Python SDK for DataMiner APIs, GQI, DOM, Automation or Data Aggregator: No plans for this at the moment.
In short, use Python for data acquisition in connectors, and use C# with AI help from the marketplace plugins for everything else. In 2027, the protobuf-based DataMiner API will open DataMiner to Python clients and AI agents alike.