Language server improvements accepted for GSoC 🎉

(Gabriel Gonzalez) #1

Folkmar Ramcke’s proposal to improve the Dhall language server implementation was accepted for Google Summer of Code. The project will be co-mentored by me and @luke. This project aims to implement a more fully featured IDE experience for Dhall, while primarily targeting VSCode as the front-end

Right now is the community bonding period, so everybody please help him feel at home here! He will be getting comfortable with the dhall-lsp-server project over the coming weeks as he ramps up. We’ll also be using this forum to coordinate so that the development process is more transparent.

(Luke Lau) #2

Congrats Folkmar! I think this will be a great project. Most configuration languages I’ve seen are very limited in what IDE support they can provide due to the lack of types and tooling, but Dhall has all this and much more information to provide a great editing experience: It’s a perfect fit for LSP.

(Folkmar Frederik Ramcke) #3

Thanks everyone! I am looking forward to getting started!

PS: I tend to go by my other first name (Frederik). Confusing, I know :wink:


Congratulations Frederik! You can ping me if you have any questions about the existing code in the dhall-lsp-server folder. Also you might want to take ownership or become a contributor of project, as it provides a VS Code plugin to go with the server. (Or not :slight_smile: On the second thought the LSP functionality is a bit orthogonal to that of the plugin.)

(Gabriel Gonzalez) #5

@panaeon: Yeah, I think we still do plan to make changes to the VSCode plugin, so we’ll need to add Frederik as a contributor. One of the things we learned while drafting the proposal was that most IDE clients still require non-trivial code to effectively use the exposed language server API. This is one of the reasons we chose to focus only on the VSCode front-end for the scope of this project.

(Fabrizio Ferrai) #6

Congrats Frederik and welcome! :blush:

(Domen Kožar) #7

Congrats to the mentors and Frederik :slight_smile:

Could I suggest considering finishing as part of LSP integration, as editor space is moving towards tree-sitter due to it’s incremental parsing attractiveness?


(Gabriel Gonzalez) #8

@domenkozar: No, that is out of scope for the project. I just created a new thread linking to the PDF of the proposed work: Proposed language server improvements