Re: [was Promotion of octave ...] Octave material for presentations

[Juan Pablo Carbajal]

On Tue, Nov 26, 2019 at 11:48 AM Kai Torben Ohlhus <address@hidden> wrote:
>
> On 11/26/19 6:45 AM, Juan Pablo Carbajal wrote:
> >
> > [...]
> >
> > Feel free to add what you consider useful. The slides you link is the
> > material I use every year at the CERN course, hence it is dynamic.
> >
> > The slides for the presentation at HSR are here [1], inparticualr look
> > at [2], all the material should be available on the root folders img/
> > and includes/
> >
> > [...]
> >
>
> Thanks, I added your material.
>
> > BTW, I remember you were working on a nb extension for publish, how is
> > it going? Also, would you be able to weigh how much work would be to
> > create a sphynx extension for Octave?
> >
> > [1] https://gitlab.com/kakila/talks
> > [2] https://gitlab.com/kakila/talks/tree/master/2018_GNUOctave_HSR
> >
>
> The efforts of this work got stalled.  Without JSON support bug #53100
> [1] this simply is too complicated parsing and writing ipynb-files
> (JSON) on my own.  Unfortunately, as time goes by the number of possible
> implementations increases and [1] got stuck in some analysis paralysis.
>
> Since I worked on publish between 2016 and 2018, Jupyter(Lab) has really
> become a great tool, which you can run offline on you local machine,
> with Octave as back-end (as I did in my presentation on Friday).  Thus I
> find there is hardly a need to further go down the road to "improve" the
> Matlab-invented publish-syntax (I prefer Markdown).

I think there is no room for discussion here. I also use jupyter(lab)
for reporting and teaching, and markdown (with math extension) is
"unschlagbar" (as of today).

>
> The remaining projects after solving [1] I have in mind is:
>
> (A) Import/open .ipynb files as Octave scripts.  All Markdown sections
> are imported as comments and remain untouched.  Output cells are ignored.
>
Is this on the Octave kernel or you would this as a function in octave?
What about a pandoc filter[1] to do this? This would be python (or
Haskel) coding, I would be interested in getting involved.

> What do you mean by "sphinx extension"?  You mean something like
>
>    publish ("myscript.m", "sphinx")
>
> resulting in some "myscript.rst" for further processing inside a sphinx
> documentation project?  This would not be too hard, just implementing
> [2] for "sphinx" or "rst".  Or being able to parse sphinx documentation
> inside m-code, similar to [3]?
> [3] https://github.com/sphinx-contrib/matlabdomain

[3] is exactly what I meant, I wonder why this did not show up in my
searches. It seems to be licensed under FreeBSD , do you see any
reasons why not to prefer sphynx over texinfo? I am pretty usre there
are opinionated positions out there for and agaisnt texinfo (I
actually never liked it too much).
Have you tested [3]?
What about texinfo <-> sphynx using sphynx? [2]

Regards,
JPi

[1]: https://pandoc.org/filters.html
[2]: https://www.gnu.org/software/texinfo/ under "Conversion to/from Texinfo"