Re: Query regarding GSOC-2020 Jupyter notebook integration project

[Kai Torben Ohlhus]

On 2/29/20 3:23 PM, Dilip Vijjapu wrote:
> Respected  sir/madam
> This is Dilip from India. I am interested in working with the Jupyter
> Notebook integration GSOC Project. I am having 2 queries regarding this
> project :
> 1. What is the ultimate outcome that your organization is expecting by
> the end of this project as there are many issues prevalent in the
> octave-Jupyter integration process
> 2. In the project description, it is written that "In combination with
> another Octave GSoC project (see JSON encoding/decoding
> <https://wiki.octave.org/Summer_of_Code_-_Getting_Started#JSON_encoding.2Fdecoding>)"
> so my query is that do we have to work on both the projects parallelly
> if we select Jupyter Notebook integration project?
> 
> So I request you to please clarify my queries so that I will get a
> crystal clear idea of what your company is expected by the end of
> the project so that I can start contributing according to the companies
> requirement.
>                                      
>                                            Thanking you


Dear Dilip,

GNU Octave is no "company".  The Octave project hopes to find in you an
interested voluntary long-term contributor to the project after having a
nice start with GSoC.  Please familiarize with the idea of GSoC and the
Octave project [1].  Short answer: it is not about doing a "job".

The first part of this GSoC project suggestion you mentioned is kept
relatively vague to improve the already existing kernel [2].  That
kernel is *not* developed by the Octave maintainers, but by Steven
Silvester.  Thus I am afraid, direct contributions to that kernel have
to happen in a fork of [2], that will eventually be merged on the behalf
of Steven Silvester.

On the other hand, I know many things that can be improved to make
"Octave a first-class experience within the Jupyter Notebook".  First of
all documentation about a proper setup is missing.  Did you ever (try
to) use (the latest) Octave within Jupyter (on Linux/MS Windows/macOS)
where you have (no) admin rights on that system?

But GSoC is not only about writing documentation, but I think it is a
good starting point to figure out what is no "first-class experience" yet.

The JSON part was added by me and is optional, if you have better ideas
what to do in your given time.  This is one larger goal I have in mind,
but requires proper JSON support of Octave, which is not given yet.

To sum up:  The project is ideal for students knowing about Octave and
Jupyter and having some ideas what they want to improve about the
experience they made with the combination of both.

Kai


[1] https://www.gnu.org/software/octave/support-expectations.html
[2] https://github.com/Calysto/octave_kernel