On Tuesday, March 31, 2020 at 7:24:52 AM UTC-4, Edward K. Ream wrote: > > Thomas, this post is an indirect response to some of your recent remarks. > There is some chance I understand them now ;-) >
It's hard to grasp some of these things until you get time to work with VR3 for a while. I have the advantage there! One reason that VR3 has good potential here is that I was interested from the start in a literate-programming style capability, to be made as easy as possible. So I was bound and determined to get the interleaved capability. > When I awoke this morning I saw that Leo + vr3 looks to be *superior *to > Jupyter! vr3 adds all essential features of Jupyter notebooks and cells, > while retaining Leo's unique features: > > 1. Interleaved @language directives is *way *better than the clumsy > dropdown menu in Jupyter that selects the default view. > It's also possible to have separate nodes for each thing, the way that Jupyter notebooks do it. With VR3, you also have the option to have more related things in the same node, which I happen to like. > 2. vr3 is aware of outline structure. > > 3. Leo + vr3 naturally retains all of Leo's superior organizational > abilities. > For me, this is *the *main reason I'd like to use Leo rather than Jupyter. And BTW, if you want to share a program with someone else, and I mean just the code itself, you can choose to render only the code blocks, export to a browser, and copy the program directly from there for pasting into a file. It would be easy to add an *Export Code to File* function, and I will probably do that at some point, but I want to keep the UI as simple as possible so I'm waiting until I get a good idea about that. > 4. In the (unlikely?) event that people want to use both Leo and jupyter, > Leo can already import/export with jupyter > Here I differ. Leo cannot import/export anything Jupyter-ish but text nodes (and maybe code, I'm not sure). Not code execution results, and certainly not live interaction outputs. That's why I posted that we would need to improve import/export to be able to properly share with non-Leo Jupyter folks. And even then, I don't see that we'd want to try to duplicate Jupyter's %-magic machinery. > *No jupyter bridge needed* > > Yes! Exactly what I've been saying. Glad you see it too! Although there > is one scenario where it might be worthwhile using a Jupyter kernel > server. That would be if you wanted to execute a very long calculation. > You'd want to do that asynchronously, so you don't tie Leo up waiting for > it to finish, and Jupyter is designed to do that. > > *Summary* > > Thomas's vr3 plugin makes every Leo node work like an *enhanced* jupyter > cell. > > All comments welcome. > > 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 [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/leo-editor/3ff9979c-e689-4fa3-b4ad-68f799290e3c%40googlegroups.com.
