All Vega rendering components are broken

[wadethestealth]

All Vega React components don't show any props inside supportedMedia, and therefore output can not be routed to their display because it is missing the default mediaType prop.

@BenRussert

[wadethestealth]

The work around for now is to downgrade to hydrogen 2.10.3 for your previously supported Vegas however it's impossible to use the newer Vegas

@phromo

[phromo]

@wadethestealth can you throw some more detailed pointers to the work needed to resolve this? In what you mention above, is that changes needed in Hydrogen or Vega/Vegalite? Maybe I could help through a PR if you point the way.

I'm currently using alt.renderers.enable('png') as a workaround to avoid downgrading.

[wadethestealth]

@phromo if you view the PR related to upgrading to support the newer Vegas #1724 there is a linked to nteract/nteract#4442 where the Vega transforms package was updated. What I know is that there were some changes there, but it could also be due to the PR that refractors our transforms #1663. (I dont believe this is related anymore as it was a part of 2.10)

[wadethestealth]

I was correct about nteract/nteract#4442. It seems that the defaultProps were left out in the refactor, so now I have fixed that, and will be awaiting the newest release of @nteract/transform-vega set to be published 9-30-19. In which case you will have all the vegas at your disposal. 😄 @phromo

[wadethestealth]

Anyone who still faces this issue should reinstall hydrogen, or if you are more familiar with npm then you need to upgrade @nteract/transform-vega to atleast 6.0.2.

cc: @phromo

[rpalloni]

Hello, does the issue still require hydrogen 2.10? I'm in hydrogen 2.14 / atom 1.44 / Ubuntu 18.04 / python 3.7 / altair 2 (or 4) and I cannot plot images with this code

import altair as alt

-- load a simple dataset as a pandas DataFrame
from vega_datasets import data
cars = data.cars()

fig = alt.Chart(cars).mark_point().encode(
x='Horsepower',
y='Miles_per_Gallon',
color='Origin',
).interactive()

fig

-- fig.to_dict() data are in

It does not seems to be an Altair issue as it works in Jupyter Notebook
vega/altair#765