It's very important that we all know how cool the valuespace plugin already 
is.Terry's recent comments became the basis for #839: Split valuespace 
plugin (or its docs) in two 
<https://github.com/leo-editor/leo-editor/issues/839>.

There are several problems with the vs- commands at present.  They can all 
be fixed in a few hours. This will happen very soon.

You could call the vs- command a kind of "Jupyter lite", allowing key 
features of IPython/Jupyter in any Leo's node. One of my long-time goals is 
to do "Jupyter heavy" (a *real* ipython/jupyter console) in selected 
(@jupyter) subtrees.

There are two essentials to move such grand visions forward:

1. The goal must be easy to state, visualize and remember.

2. Short-term (1 hour to 1 week) actions must present themselves.

Both have just happened!

*Goal*:  A Jupyter console in the body pane.

*Aha*: True, fully integrated, support for Jupyter will require a *Jupyter 
gui plugin*.Indeed, Jupyter's existing qtconsole code is based on a Qt 
QMainWindow object. A lot of work will be needed to "rewrap" all that code 
for use in Leo's code base. This might take as long as the console gui 
plugin, but so what? The payoff will be *hugely *greater. Leo could easily 
become a killer Jupyter app!

*Actions*: I am now studying the Jupyter sources intensely, and have 
written prototype code that will launch a Jupyter app in Leo' console.  The 
prototype isn't remotely ready for prime time, but it allows me to inspect 
live code.

The Jupyter/IPython code base is huge, and contains lots of wrapper code.  
This is inevitable and can't be changed. jupyter_console.ptshell contains a 
lot of "meat" that may inform our thinking of vs-eval-block. In particular, 
the handle_image* methods do a lot of image-related magic behind the 
scenes. Imo, this magic is simpler for Jupyter users than the VR plugin.

*Summary*

An improved valuespace plugin (#839 
<https://github.com/leo-editor/leo-editor/issues/839>, with docs that don't 
bury the lede) will *quickly *bring many of Jupyter's benefits to Leo.

Close study of the Jupyter code base shows how much work Jupyter does 
behind the scenes and suggests improvements to VR and valuespace plugins.

Support for Jupyter will, in effect, require a new gui plugin.  But this 
project is worth any amount of work, and the "next first actions" are 
within reach. I am excited about the possibilities.

Edward

-- 
You received this message because you are subscribed to the Google Groups 
"leo-editor" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to leo-editor+unsubscr...@googlegroups.com.
To post to this group, send email to leo-editor@googlegroups.com.
Visit this group at https://groups.google.com/group/leo-editor.
For more options, visit https://groups.google.com/d/optout.

Reply via email to