table.show_in_notebook hangs for modestly sized table (~4500 rows) #11836

Description

I attempted to view an astropy table using the show_in_notebook() feature, but it caused my Jupyter notebook to hang/crash, instead. I don't know what's causing the task to crash; I would not normally expect a ~4500-row table to exceed memory limits for something like this, but perhaps it is too large of a table for this feature? (If so, I'd expect a warning or exception, rather than the browser locking up.)

I investigated a little further, and found that I can explore the performance more carefully by doing table[0:100].show_in_browser(), and gradually increasing the number of rows included. On my system, for this file, show_in_notebook() has the following performance for the following table subset lengths (timings estimated using phone stopwatch app):

1. 200 rows: below ~1s
2. 300 rows: ~2s
3. 400 rows: ~3s
4. 500 rows: ~4s
5. 750 rows: ~11s
6. 1000 rows: ~25s
7. 1500 rows: 1m15s
8. 2000 rows: 2m18s

I don't have the patience to go further, but the trend is clearly superlinear and gets really sluggish to the point of uselesness beyond 500-1000 rows.

The following two tables from the literature (Robberto et al. 2020, for those curious) each cause this crash for me:

Gists for Tables 2 and 4
Table downloads: Table 2 | Table 4

Expected behavior

I expected that the table would be shown using the show_in_notebook viewer.

Actual behavior

The tab crashed/hanged, and I waited for several minutes - nothing changed, I got a "loading wheel", and my other Jupyter tabs also hanged.

Steps to Reproduce

1. Load the table using a step like this: robb_table2 = astropy.table.Table.read(filepath_table2, format='ascii')
2. View the table using robb_table2.show_in_notebook()
3. The tab hangs.

# Put your Python code snippet here.

System Details

Darwin-19.6.0-x86_64-i386-64bit
Python 3.7.7 (default, Mar 26 2020, 10:32:53)
[Clang 4.0.1 (tags/RELEASE_401/final)]
Numpy 1.20.2
astropy 4.2.1
Scipy 1.6.2
Matplotlib 3.3.4

I'm on Mac OS X 10.15.7 Catalina, and Firefox 89.0, if that matters.