Update R version to 3.3* in r-notebook and datascience-notebook

[poplav]

No description provided.

[parente]

Thanks @poplav . I'm curious to see how many of the packages work with R 3.3. :)

[poplav]

Just realized there is the datascience-notebook as well [hand over face]. I'll update there as well.

[parente]

@poplav No worries. I clearly don't have all the places nailed down in my mind either!

[poplav]

I am seeing #210. I need to verify it is not a result of these changes, so hold off on merging please.

[poplav]

It looks like 2485724a08c9 reintroduced #210 making libjpeg v9. Noting this is then not attributed to the recent R upgrades.

[parente]

Thanks for tracking that down. I'm going to do one last check on the all-spark-notebook vs datascience-notebook for the tag you mention. I'm wondering if the R problem only exists in one and not the other because of the libs that are touched by other conda actions.

[poplav]

I did see that the fix mentioned in #210:

echo "jpeg 8*" >> /opt/conda/conda-meta/pinned
conda update --all -y

does work after the image is built. In my case I could run this in the dev_image step of declarative widgets, but that would not be optimal/time consuming

[parente]

Nah. Don't fix it in your build. I'm looking into how to fix it in the main images.

[parente]

Pushed 97a5071c5775.