Hi Serge,
On 29/09/15 06:43, Serge Stinckwich wrote:
Looks interesting !
Thanks :).
Did you have a look to Jupyter notebook ?
https://jupyter.org/
Yep, I know it. In fact I come from IPython to Pharo because I need a
more moldable and flexible environment for data narratives and
visualization and adapting IPython/Jupyter has a lot of cognitive burden
when is not development but writing/research your primary goal. I have
wrote about it here[1][2]:
[1]
http://mutabit.com/offray/static/blog/output/posts/grafoscopio-idea-and-initial-progress.html
[2]
http://mutabit.com/offray/static/blog/output/posts/on-deepness-and-complexity-of-ipython-documents.html
Thinking now about it, it occurs to me that grafoscopio is to Jupyter
kind of what SmalltalkHub is to GitHub. It tries to solve the same
problem of Jupyter (interactive documentation) but with a different
approach, from a "pharo/smalltalk" perspective. For me the big advantage
are tree like interactive documents (as pointed in [2]), a feature that
Jupyter notebooks are still lacking and that starts to weight when you
start to write a long interactive document on Jupyter with only cells.
You can split the document on several files, but you lost panoramic view
(you see only the current file) or have a single long one file, but then
your are lost in the details (see an example of an abandoned notebook
exploration at [3])
[3] http://mutabit.com/repos.fossil/piamed/doc/tip/Afiche/narrativa.png
This could be nice if we could have a Pharo support for this (already
supported by other
languages like R, Python, Scala, etc ...).
Best,
Or have it the other way around: R, Python, Scala supported in Pharo,
without the constrains of a web interface and the notebook "cells only"
format. Maybe some kind of support for the ZeroMQ jupyter kernel could
be the start of this kind of multi-language support *inside* Pharo
documents.
Cheers,
Offray
