Hi, first post here.
I am a Julia & Jupyter user, and have run into a problem that makes my workflow very problematic. The issue was originally posted in the vscode github, because it was discovered in the vscode-jupyter environment. But since then we found out it is also present in plain jupyter-in-browser.
Problem description: displaying a svg plot produced with VegaLite takes a very long time that scales linearly with the number of elements in the plot. The slowness is only seen under macos in any browser I tested. It occurs both in a classic notebook ui and in jupyterlab. Under Windows there is no issue, the display is fast and agile. There is also no issue in macos if the plot is displayed in the browser without using jupyter (e.g. launching a browser tab with the plot from the Julia REPL). Macos version is 11.6.1.
Here is a minimal Julia code that produces the issue
using VegaLite
npoints = 10000
@time [(; x = rand(), y = rand(), z = rand()) for _ in 1:npoints] |> @vlplot(:point, x=:x, y=:y) |> display
This particular plot with npoints = 10^4 points takes 50 secs to display on my macbook, or around 5 secs with npoints = 10^3 points.
Before opening a github issue I wanted to ask here if this is a known issue.