Add medium_cells benchmark #41
Conversation
This patch adds a new benchmark which aims at filling a gap between the `dense_cells` and the `light_cells` benchmarks, providing a more realistic example of escape sequence density in a normal terminal workflow. The new benchmark is an Alacritty ref test recording of a short NeoVim session, which includes several small and large buffer movements, different styles of underlines, and window-in-window using a fuzzy finder and tiled buffers. This new test intentionally uses a workflow that is pretty escape sequence heavy, while still being realistic, since workflows without heavy escape sequence use are already reasonably well covered by the `light_cells` benchmark. Closes alacritty#40.
|
Running this now leaves some junk in my prompt afterwards. Tested using also noticeable with I'm just mentioning it here in case someone else notices this too. I don't think it's anything wrong in the benchmarks, but an unresolved issue in alacritty that I've lost track of. |
It's not really an unresolved issue in any way. This is just the terminal responding to some of the queries made by NeoVim during the benchmark. |
|
Just to follow-up on your question @nixpulvis, since I've looked into possible ways to improve this: The issue here is obviously that Alacritty on itself is writing replies to the terminal asynchronously. This is particularly problematic because the application itself could already be shut down by the time the reply arrives. I played around a bit with trying to non-blockingly drain STDIN, but everything I've tried has just been a giant mess and only worked relatively unreliably (at least without timeouts). So I think it's easier to just consider the terminal the benchmark is executed in somewhat sacrificial, or at least having to do some minor 'cleanup' after the run completes. |
This patch adds a new benchmark which aims at filling a gap between the
dense_cellsand thelight_cellsbenchmarks, providing a more realistic example of escape sequence density in a normal terminal workflow.The new benchmark is an Alacritty ref test recording of a short NeoVim session, which includes several small and large buffer movements, different styles of underlines, and window-in-window using a fuzzy finder and tiled buffers.
This new test intentionally uses a workflow that is pretty escape sequence heavy, while still being realistic, since workflows without heavy escape sequence use are already reasonably well covered by the
light_cellsbenchmark.Closes #40.
Runtime locally seems slightly above
light_cells(~9ms) at ~13ms.