Gregory Anders

Fixed by #28676.

> I do wonder if the treesitter highlighter wants to set, `flush == true` then. I think this was an improvement: [#28101 (comment)](https://github.com/neovim/neovim/pull/28101#issuecomment-2091057965). Hm if I set `flush = true`...

#28676 does indeed fix the problem this was intended to solve, so I leave it up to others to decide the fate of this PR (I am fine with either...

This looks similar to https://github.com/neovim/neovim/issues/26372 which was fixed in https://github.com/neovim/neovim/pull/26381. I can't reproduce this on https://github.com/neovim/neovim/commit/8bb5089974e14f043a0b0c17e4d38dbd0b9a9ab2 but I can on current master, so I think this might have regressed.

Bisected to 037ea6e786b5d05f4a8965e4c2ba6aa60ec7c01a. cc @luukvbaal **EDIT:** Hmm there might be more than one thing going on here. Using this as init.lua: ```lua vim.api.nvim_create_autocmd("VimEnter", { callback = function() --[[ (1) ]]...

> I don't think you are bisecting correctly. The previous commit has the same behavior. I am bisecting correctly. Please read my edit.

I've found a separate but similar issue, which I at first thought was the same. I'll open a separate issue for this.

`vim.Diagnostic` is not an "LSP diagnostic" though, and we try very hard to avoid imposing LSP specific concepts into the (generic) diagnostic structure. In this case it is probably fine...

Can you please also update `news.txt` and add a test case to `editorconfig_spec.lua`?

> > plugins are not able to detect if an existing mapping is one created by the user or one created by Nvim. > > The reason for choosing to...