Version on binder does not plot.

[koldunovn]

Currently, the version that is opened with the binder does not make the plot, even if I click the Plot button.

[martindurant]

I believe this is the same as #69
I will try to make some time to put things right.

[martindurant]

Works OK now?

[rsignell-usgs]

I clicked the binder link, ran the notebook, got a plot, but then (yipes) I get this:
https://www.youtube.com/watch?v=50dylR5d3-M

[martindurant]

Constant unzooming! Is there an issue for this?

[rsignell-usgs]

Do you think it's a holoviews/geoviews issue or an xrviz issue?

[martindurant]

As far as I know, xrviz creates the plot and then leaves it alone. It is plausible that the layout is being recalculated somehow... @philippjfr , your thoughts?

[philippjfr]

Yeah something in holoviews. It's probably to do with aspect handling. Is this using recent HoloViews yet?

[martindurant]

The script uses "current" versions (whatever conda gives us from conda-forge on py37)

[martindurant]

@rsignell-usgs , if you still have this open, would be good to post the versions binder ended up using

[rsignell-usgs]

# packages in environment at /srv/conda/envs/notebook:
#
# Name                    Version                   Build  Channel

bokeh                     2.0.1            py37hc8dfbb8_0    conda-forge
datashader                0.10.0                     py_0    conda-forge
holoviews                 1.13.2             pyh9f0ad1d_0    conda-forge
hvplot                    0.5.2                      py_0    conda-forge
geoviews                  1.8.1                      py_0    conda-forge
geoviews-core             1.8.1                      py_0    conda-forge
panel                     0.9.5                      py_1    conda-forge
python                    3.7.6           h8356626_5_cpython    conda-forge

[philippjfr]

Okay, so definitely latest, I'll try to reproduce.

[philippjfr]

Okay, I see the same thing. Will see if I can get a 1.13.3 release with a fix out very soon.

[koldunovn]

The plot works for me now on the binder, thanks!

[rsignell-usgs]

@philippjfr I'm so glad you could replicate the zooming out feedback loop issue and it sounds like you see the solution also. Awesome!

[philippjfr]

Just wish we had an automated way to catch these kinds of regressions. Unfortunately testing client-server interactions in a browser is hard.