Skip to content

Conversation

@rtfeldman
Copy link
Contributor

We decided to take this out for now. It doesn't seem necessary, and it complicates the code a lot. We can always put it back later if desired.

Release Notes:

  • N/A

Prompt was:

> remove crates/scripting_tool and update all code paths that were using it to no longer do that. keep fixing diagnostics until they are all fixed.
@cla-bot cla-bot bot added the cla-signed The user has signed the Contributor License Agreement label Mar 24, 2025
@maxdeviant maxdeviant merged commit 5d05c4a into main Mar 24, 2025
15 checks passed
@maxdeviant maxdeviant deleted the remove-lua-tool branch March 24, 2025 19:58
@mocenigo
Copy link

mocenigo commented Sep 16, 2025

As I wrote here: #38249 (comment)

Currently Zed has several files to implement various LLM agents, just have a look at crates/agent_servers/src.
This implies that if a change is done to the protocols, or some features are added to the various providers, the corresponding files need to be modified and the entire application updated. In fact, this also requires an application restart.

If these were all loadable extensions, maybe even written in a scripting language (such as Lua, or even the mini-R7RS scheme like stak), then updates would be much faster, and support for LLM providers could be fixed even without having to restart the application.
I understand that this would be too much before 1.0: the extensions API would need a significant expansion in any case, a scripting language should be added if we do not use just WASM-compiled extensions, etc. But I suggest that it should be considered post-1.0.

I believe Lua would be ideal for this. There would be no need to rewrite everything. Only new stuff, whereas the fact that LLM provider support is placed in extensions written in Rust would make updating them easier anyway.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

cla-signed The user has signed the Contributor License Agreement

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants