Post

This is some great benchmarking. Honestly, even though there is a statistically significant difference in startup times, 2-3ms is not noticeable to an end user. With all the supply chain attacks going on, I’d be more worried about the risk of running 50 different plugins that could be compromised.

Hey Evgeni! Thank you for the article. I'm curious about this bit: > consider doing some simple lazy loading (like after first screen draw or on InsertEnter/CmdlineEnter events) I understand the event part: you use autocmds for those. But how do you run something "after first screen redraw"?

The way I do it is with `vim.schedule(function() ... end)`. When inside config file, it will be executed after first draw. A more convenient solution is to use `later()` from 'mini.deps'. I also plan to add a more involved function to 'mini.misc', since 'mini.deps' will retire eventually.

Cool! Not too bad of a friction. Too bad I'm not using Quarto for my blog (using Minimal Mistakes theme via GitHub Pages, I'm a simple man)

Great post! Now I need to decide if I want to save those extra ms and just use mini.nvim :) Currently, I'm using a total of 57 plugins 😳

Since then, I refactor my entire config and it is now 17 plugins and less than 400 LOC (single file for now). Using the new vim.pack :)

I will consider it :) [I gave my plugin-list to Claude and it claims that there are indeed 8 of them which have a mini.vim equivalent or a close alternative]