Support processes in rendering HoloViews

[ahuang11]

I think processes=True (the default) now works @rsignell

import xarray as xr
import hvplot.xarray
from streamjoy import stream

ds = xr.tutorial.open_dataset("air_temperature").isel(time=slice(1000))
stream(ds.hvplot.image("lon", "lat"), uri="air.html")

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 83.33333% with 7 lines in your changes missing coverage. Please review.

Project coverage is 69.90%. Comparing base (c253785) to head (bf84f58).
Report is 5 commits behind head on main.

Files	Patch %	Lines
streamjoy/_utils.py	83.33%	4 Missing ⚠️
streamjoy/serializers.py	70.00%	3 Missing ⚠️

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #47      +/-   ##
==========================================
- Coverage   69.95%   69.90%   -0.06%     
==========================================
  Files          14       14              
  Lines        1448     1462      +14     
==========================================
+ Hits         1013     1022       +9     
- Misses        435      440       +5     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[ahuang11]

Unfortunately, it's still slow. I think I need to cache the webdriver instance per worker somehow.

[ahuang11]

If I can figure this out...
https://dask.discourse.group/t/how-to-cache-an-object-instance-per-worker/2801

[ahuang11]

okay :D I figured it out with a lock and caching it in get_worker.

I think it takes 1/3 the time of vanilla HoloViews rendering GIF time.

[ahuang11]

Found the problem; I didn't set:
hv.Store.save_option_state = True

holoviz/holoviews#6286