Re: [Maxima-discuss] Maxima-Jupyter

[Fraser D. <dos...@ac...>]

May have been a permissions issue - cloning from git again and making 
sure I own the files

On 2020-10-16 2:00 p.m., Robert Dodier wrote:
> On Fri, Oct 16, 2020 at 8:48 AM Fraser Doswell <dos...@ac...> wrote:
>
>> Has anyone succeeded in building and using the maxima-jupyter kernel?
>>
>> I tried on ubuntu 20.04 but my maxima was built with Gnu Common Lisp,
>> not SBCL.
> Maxima-Jupyter requires a threading library (Bordeaux threads) which
> isn't supported by GCL, so it *cannot* work with GCL.
>
> Maxima-Jupyter is known to work with SBCL and Clozure CL; it might
> work with other Lisps which support Bordeaux threads, but my advice is
> to stick to SBCL or CCL.
>
> For SBCL, and probably for CCL as well, it's necessary to build Maxima
> as a so-called loadable core image. The Lisp runtime is executed and
> then it loads the core image. This makes available additional stuff
> (notably the package manager ASDF) which is needed for Maxima-Jupyter.
> The core image is in contrast to a so-called executable image, for
> which the Lisp runtime isn't needed, and therefore ASDF etc. are not
> necessarily present.
>
> For Linux systems such as Ubuntu, it's straightforward to install
> SBCL, build Maxima, and then build Maxima-Jupyter. I should really fix
> up the installation readme, but let me outline a general strategy and
> we can go into details as you progress through it.
>
> (1) install Python (probably python3)
> (2) install Jupyter
>
> At this point you can try the existing Jupyter kernels and just get a
> feel for how Jupyter works in general. It handles all the stuff around
> the edges which is super helpful, so when you switch from one kernel
> to another it still works largely the same. List of existing kernels:
> https://github.com/jupyter/jupyter/wiki/Jupyter-kernels There are a
> lot of example notebooks there too.
>
> (3) install SBCL. I don't remember for sure if there is an installable
> binary package; I guess I hope so, but if not, you'll have to build it
> from source.
> (4) build Maxima. This is the same workflow as for many other programs:
>
> download maxima-5.nnn.tar.gz from
> https://sourceforge.net/projects/maxima/files/Maxima-source/
> $ tar xvzf maxima-5.nnn.tar.gz
> $ cd maxima-5.nnn
> $ ./configure --enable-sbcl
> $ make
> $ sudo make install
>
> (5) build and install Maxima-Jupyter. Before going into that, how
> about if you try the preceding steps and let us know how it's going.
> I'll try to clean up the readme so that it's more useful.
>
> (6) launch Jupyter and select the Maxima kernel. You're in business!
>
> OK, this is just a sketch. We can work through details as things come
> up. Good luck and have fun!
>
> best,
>
> Robert Dodier