From c080c8a8ed51ca15b9d681faa94f6a23a481473b Mon Sep 17 00:00:00 2001
From: atinylittleshell <3233006+atinylittleshell@users.noreply.github.com>
Date: Mon, 15 Jan 2024 12:55:38 -0800
Subject: [PATCH] Create PERFORMANCE.md

---
 PERFORMANCE.md | 11 +++++++++++
 1 file changed, 11 insertions(+)
 create mode 100644 PERFORMANCE.md

diff --git a/PERFORMANCE.md b/PERFORMANCE.md
new file mode 100644
index 0000000..b4fd22e
--- /dev/null
+++ b/PERFORMANCE.md
@@ -0,0 +1,11 @@
+Benchmarking the performance of terminal emulators is tricky. There are tools such as https://github.com/alacritty/vtebench but as called out by the author it's not good representation of the actual E2E experience, and the best way is to simple test specific use cases.
+
+while I continue to look for more systematic approaches for benchmarking, here's a video looking at smooth scrolling in neovim as a key use case TerminalOne optimizes for: 
+
+- Scrolling through the same file with the same Neovim setup on the same PC (11th Gen Intel i9 3.50GHz + GeForce RTX 3080 Ti)
+- First half is WezTerm (No criticism intended! WezTerm is an awesome terminal. Using it for comparison since it's commonly raised as a target for benchmarking)
+- Second half is TerminalOne
+- This comparison is mainly to highlight that Javascript doesn't inherently create performance bottlenecks - moer often than not they are created by the application layer not the language
+
+https://github.com/atinylittleshell/TerminalOne/assets/3233006/6f7dee53-7c8b-4991-9bdc-3250547a3945
+