You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Since GHC 8.8+ can now generate Language Server Index Format files via hie-lsif, lsp-haskell supporting them would allow performant lsp features without having to install language servers or having potentially slow language servers running.
This is especially great since Haskell IDE Engine has had performance issues according to many and in my own experience trying to use it. Though there are work arounds such as killing it when it goes over 2GB of memory, that can cause system slowdowns that are a liability to workday productivity.
I believe this is the proper place for supporting these files in emacs, let me know if not.
I found some related discussion in neoclide/coc.nvim#993. To prevent any information not being relayed or lost, I also put out a question on the lsp-mode gitter.im here
I think this probably needs to be supported in lsp-mode. lsp-haskell is just a thin wrapper that tells lsp-mode how to launch a server.
Supporting lsif files should (I believe) be done in a language-agnostic way. When/if lsp-mode supports them then we would need to do whatever work thy require to generate them.
Since GHC 8.8+ can now generate Language Server Index Format files via hie-lsif, lsp-haskell supporting them would allow performant lsp features without having to install language servers or having potentially slow language servers running.
This is especially great since Haskell IDE Engine has had performance issues according to many and in my own experience trying to use it. Though there are work arounds such as killing it when it goes over 2GB of memory, that can cause system slowdowns that are a liability to workday productivity.
I believe this is the proper place for supporting these files in emacs, let me know if not.
As a starting point, I guess looking at the vscode blog post for lsif, the LSIF specification, and potentially the vscode-lsif-extension forked to add Haskell support by @mpickering.
Any direction in implementing this or what would be involved in implementing it is much appreciated.
The text was updated successfully, but these errors were encountered: