วีดีโอ

Thank you so much for the kind words.

Good luck and have fun!

You're welcome. I'll think about that one. Stay tuned!

Right on, congrats! And I appreciate the heads-up about attaching.

Thanks for watching. I might revisit it at some point 😀

I'm just a dude. Thanks for watching.

You’re welcome! Thanks for watching

Will do. Thank you so much for the kind words

Glad to hear that!

Thank you for the kind words. This is a good question. There's a few things that can go wrong here. First I'd check to see if running the `cmd` for the `vim.lsp.start` runs fine when you run it directly from the terminal (it shouldn't do anything, but it should sit there waiting for input without crashing). If that works, then your neovim probably doesn't have the same $PATH setup as your terminal and you can pull on that thread (or specify absolute paths, etc). If that didn't work, then hopefully it exited with a helpful error message. Let me know how it goes!

@@semanticartThanks for the quick reply! I confirmed that running the command "npx ts-node <absolute/path/to/server.ts>" did kick off the server. I am on Windows and have never used neovim before so it may be something with my neovim config or something. I didn't mention at first but the error message also includes "Spawning language server with cmd: {"npx", "ts-node", "<absolute/path/to/server.ts>" failed." So it is successfully trying to spawn it with (seemingly) the right command. I don't know where to look for more details on the failure. (like a neovim error log file or something?)

@@PriZ0nM1ke try this. Hit colon to get into command mode and do !npx ts-node <absolute/path/to/server.ts> and hit enter. See if you get any helpful message from that.

@@semanticart It runs the file (I put a console.log() at the start to confirm) but doesn't hang like in a terminal. Doesn't throw any errors as far as I can tell.

@@PriZ0nM1ke Hrm. Check the lsp log file (get the name via `:lua =require('vim.lsp.log').get_filename()`). You can also set `vim.lsp.set_log_level("DEBUG") ` BEFORE starting the language server to (hopefully) get more info

No. You can write one in literally anything. Here's a very boring example I wrote in bash: prefab.cloud/blog/writing-a-language-server-in-bash/

Thank you so much for the kind words! Best of luck with your language server

Monaco :) I've tried so many fonts but always come back to Monaco

Not sure if you're trolling or not, but this video has nothing to do with Golang. Also, there are language servers for Java, C#, and PHP.

That sounds fun! Let me know how it goes. I don't know when I'd be able to make a video about that, but I can give you a quick overview of how you could do it. 1. As you've seen in the videos, you need to keep track of the files that are open in the editor. You can do this by listening to the `textDocument/didOpen` and `textDocument/didClose` notifications. 2. When you hover over a symbol, you need to find the file that contains the definition of that symbol. You'd need to do this by parsing the imports of the file that you're currently in. 3. If the file that contains the definition of the symbol is open, you can show the value of the symbol. If it's not open, you can read the file from disk and show the value.

@@semanticart appreciate the response! That sounds like the next big step. I spent all day Saturday and Sunday messing with it. I got a lot of it working! I’ve got hover and completion working (for this file only, it doesn’t look at other files yet). There’s an added layer of difficulty because for the language I’m doing the #includes are not direct path names. I have to find it through a database. I’ve been working on trying to right an EBNF into an ANTLR. Then I’m going to attempt to make a parser so I can tell the user about syntax errors. I’m not even sure if that’s a thing/ the thing to do but yeah. Any quick tips about showing syntax errors?

@jondanford That's some great progress! I expect that diagnostics should cover you on syntax errors. You can specify the range of the document they cover, severity, and the message content. microsoft.github.io/language-server-protocol/specifications/lsp/3.17/specification/#diagnostic

@@semanticart Yeah! That is what I am currently using, but I was hard programming the syntax errors myself - like a noob. I am trying to add a parser that will return errors that I can then send back with diagnostics. (Hopefully) I think my grammar has some errors in it, but that can be ironed out later.

@@jondanford Gotcha. If there's an existing compiler/parser that returns errors, I'd try (ahem) parsing those from your LSP. If not , then godspeed working on your parser/grammar 🍀

Thanks for watching!

Great question! For the basics, absolutely. I have an old blog post here about using Pandoc to process the markdown for this purpose blog.semanticart.com/2014/01/11/writing-hypertext-fiction-in-markdown/ What's missing is that Pandoc's error messages are tailored to generating valid html/etc., not to helping the user craft a valid story. My goal is to build an entire suite of tools here under the MIT license that anyone can use as they wish. Stay tuned for more!

@@semanticart nice, thanks for the blog.

Thanks for watching and for the kind words. I'm glad you're able to follow along well and that the speed works for you :) Apologies for any confusion related to capabilities. Editing is hard and I'm still figuring it out 😅

That's awesome! LMK if you end up publishing that LSP, I'd love to take a look. I'm getting into Go, so I'm sure to have some future LSP+Go content here.

Glad I could help! If you publish your LSP somewhere, please send me a link :)

My pleasure! I'm glad you enjoyed it :)

That's a fair point. Thanks for the comment!

Glad it helped!

Rishi, it was great chatting with you today. Best of luck with your project!

Thank you. It is possible but it is trickier and depends on the execution context of the editor, etc. It is outside my area of expertise but I know that twitter.com/PaoloRicciuti has done this for www.sveltelab.dev/ with some combination of Codemirror and/or webcontainers.io/ He's a nice chap so I'm sure he'd be happy to give you a few pointers.

Glad you liked it!

If you're implementing functionality covered by the LSP, then you shouldn't have to implement any functionality in the IDE (if the IDE/editor already supports the LSP -- many do). You should only have to implement the functionality in your language server and all LSP-enabled editors can use it for free. If you're implementing custom functionality, then yes, you'll need to implement things in both your language server and some glue code per IDE. That's non-trivial for sure, but putting the functionality in a language server means you get to re-use that for every editor and just have to handle the glue code for the editors. This is much less work than implementing ALL the functionality (both the UI and backend code) in each editor. :) I have a video where I look at a ruby-lsp custom method that only has support in VS Code and implement the glue code for neovim. You can check it out here: th-cam.com/video/8_u0Fvtq_ew/w-d-xo.html

Glad it helped!

Thanks! Have a great day

Thank you for the kind words and have fun following along 😀

I hope you enjoy it! My PR was merged so you can use the official docs to get this feature :) github.com/Shopify/ruby-lsp/blob/main/EDITORS.md

I like this. It makes sense especially when you're trying to document the broadest usage of a tool. e.g. I could imagine a page for `pandoc` where 1. a sentence describing pandoc 2. ` brew install pandoc` to show how to install it (simplified because we've detected you're on macOS) 3. `pandoc example.md -o example.html` to show an example usage I'll think more about this. Thanks for the suggestion!

And thank you for the great idea! I have a feeling cognitive breakpoints will soon be *the* way to write documentation. I, for one, will definitely be using it. And maybe even add a global and local preference, like: 2 would be my preferred breakpoint. Then, while scrolling down, I might hit a section that I don't quite understand, so I can change the breakpoint only for said section without affecting the rest (just another thing to think about).

​@@D33P-5H4D3​ I love the idea of site-wide preferences. I've also been thinking about per-section sliders too for expanding (e.g.) just code example. Per-section sliders feel overwhelming, but time will tell. Please LMK if you do anything with this in the wild 🙌

Thank you for the kind words! I hadn't thought about trying to CSS-ninja my way through this but you're probably right that one could.

🤔I'm not sure what C4 is in this context? Mind sharing a link?

Yay! Please lmk if you use some variation of this. I’d love to see it in practice. I’m super down to collab if I could help!

Thank you, kind sir!

Thank you for the kind words

Thanks, Kassio! This is a bit tricky. The ruby-lsp-rspec addon exposes CodeLenses for Run, Run In Terminal, and Debug. I suspect those should show up in your Neovim (assuming you have a current-ish Neovim) but there might be some intelligent feature-detection to avoid showing these in Neovim (hard to say without doing some digging). But then the problem is that Neovim doesn't know how to handle the command associated with each CodeLens. The VS Code extension specifies the commands (example: github.com/Shopify/vscode-ruby-lsp/blob/b8771484173f8fe9324e6a14f5d16e63b55686ee/package.json#L70-L73 ) and then later registers the commands (example: github.com/Shopify/vscode-ruby-lsp/blob/b8771484173f8fe9324e6a14f5d16e63b55686ee/package.json#L70-L73 ) To make this work in Neovim, you'd need to add handlers for each of these commands you wanted to support. The implementation would depend on what test runner you're using in neovim. I'll definitely consider exploring this and making a video. How do you currently run tests in Neovim?

You're super welcome, thanks for watching. It means a lot to hear about real-world impact. You've got a new subscriber :)

Thanks! I'm glad you're enjoying them.

😄 Cheers!

This comment means a ton to me. Thank you so much for your kind words and encouragement. I’ll try not to disappoint 😀

PR is here: github.com/Shopify/ruby-lsp/pull/1389 ...and it is merged!

Better Finish than unfinished 🥁

Thanks. I use bun's built-in testing on other projects, I'm just trying to keep keep things more vanilla here :)

Absolutely Jeff. Please keep them coming. Would love to see how these language servers can work in Monaco in the web. Thank you so much.