GitHub user AlexStocks added a comment to the discussion: [Proposal] Pixiu AI 
evolution should focus on Dubbo-native differentiation

I think there is also an important **joint strategy** question here:

> How should `dubbo-go` and `pixiu` work together in the AI era, instead of 
> evolving as two disconnected AI narratives?

My current view is:

- `dubbo-go` should primarily evolve as the **AI-ready service substrate**
- `pixiu` should primarily evolve as the **AI capability gateway / control 
plane**

That means the two projects do **not** need to compete on the same layer.
They can form a cleaner division of labor:

## Suggested division of labor

### `dubbo-go`
Focus on service-side capabilities such as:
- AI-friendly service metadata
- streaming/session semantics
- tool/schema exposure primitives
- AI-specific observability on the service/runtime side
- governance primitives inside service invocation and service discovery

### `pixiu`
Focus on entry/control-plane capabilities such as:
- service-to-tool exposure
- model/tool/API traffic routing
- policy enforcement
- tenant/cost/security controls
- MCP / gateway / ingress integration
- end-to-end observability from user request to downstream service/tool result

## Why this split makes sense

If both projects try to become “AI platforms”, they will overlap and dilute 
each other.

If they instead form a layered story, the overall Dubbo ecosystem becomes much 
clearer:

- `dubbo-go` = where enterprise service capability is produced and described
- `pixiu` = where enterprise service capability is exposed, routed, secured, 
and governed for AI/API consumers

In other words:

- `dubbo-go` answers: **how should services become AI-ready?**
- `pixiu` answers: **how should AI systems safely consume those services at 
scale?**

That seems much more defensible than having both projects independently add 
generic AI features.

## Possible follow-up RFC split

If this discussion gains traction, I think it can be split into a few narrower 
topics:

1. **RFC: Dubbo service metadata model for AI/tool exposure**
   - target mostly `dubbo-go`

2. **RFC: Pixiu service-to-tool gateway model**
   - target mostly `pixiu`

3. **RFC: Registry/metadata-driven AI routing across dubbo-go + pixiu**
   - joint topic

4. **RFC: End-to-end observability for model → tool → downstream service**
   - joint topic

5. **RFC: Security/governance boundaries between service substrate and gateway 
control plane**
   - joint topic

I think clarifying this ecosystem-level cooperation model would make the Pixiu 
AI story much stronger than discussing Pixiu in isolation.


GitHub link: 
https://github.com/apache/dubbo-go-pixiu/discussions/990#discussioncomment-17431513

----
This is an automatically sent email for [email protected].
To unsubscribe, please send an email to: 
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to