TW on Twitter -- Computational Notebooks in TW? ...

[@TiddlyTweeter]

This didn't go publicly far yet because TiddlyWiki, though slowly growing on Twitter, doesn't yet have quite enough followers. (I'm sure it will come. Its grown a lot over the last year.)

But this was a very interesting recent conversation about "Computational Notebooks".

I not seen anything about them here so thought I'd just post a bit of it. Here is just the last part of it ...

Greg Wilson @gvwilson
I keep coming back to TiddlyWiki tiddlywiki.com and en.wikipedia.org/wiki/TiddlyWiki

Greg Wilson @gvwilson
What would a computational notebook based on TiddlyWiki look like? Is there a grad student out there somewhere exploring this space?

Mike Stok @MikeStok
Replying to @gvwilson
@joeerl might have an idea of what a computational notebook based on TiddlyWiki might look like… maybe

@TiddlyTweeter
Replying to @MikeStok @gvwilson @joeerl
... also ask @EvanBalster

Evan @EvanBalster
Replying to @TiddlyTweeter @MikeStok and 2 others
Super interested in this topic. I implemented recursive spreadsheet math for TW to probe this space. evanbalster.com/tiddlywiki/for…

Greg Wilson @gvwilson
Replying to @EvanBalster @TiddlyTweeter and 2 others
I'd be very interested in chatting more - I'm gvwilson@******.com if you have time to do so. Thanks!

[@TiddlyTweeter]

If you wondering what a "Computational Notebook" is and what the current enthusiasm about them is about, read ...

1 - https://en.wikipedia.org/wiki/Notebook_interface ... which shows the idea is not new, and immediately you can see that TW can be that ...

2 - https://hal.archives-ouvertes.fr/hal-01676633/document ... a very good academic text about them.

[Andreas Hahn]

Well,
In my opinion TW is ill-suited as a computational notebook and the
already established solution in the Form of Jupyter is more than up to
the task. The only think I can see happen is TW having an interface to
the Jupyter Kernels, but even then, it doesn't quite fit TW's model.
The main problems seem to be:

- TW has a monolithic design, whereas for a computational notebook, the
Jupyter kernel <-> Interface divide serves it very well as you can
detach computation from the code/notes and therefore scale things much
better and work with different languages. For TW all I could think of
here would be WebWorkers.

- TW has a limit on data size. Nowadays, most datasets, no matter how
small are already larger than what TW might be able to handle, so they
would have to be in an external file/files either way.

So while I can see it being nice to have TW as an interface where your
results get stored and commands are issued, I don't quite see it play
out any of its advantages in this field and think TW's architecture puts
both hurdles and limitations on what you would be able to do.

/Andreas

[Mark S.]

"It pairs the functionality of word processing software with both the shell and kernel of that notebook's programming language."

That sounds a lot like TW. TW would need a math plugin to easily get to the "computational" part. Possibly a platform for students or fieldworkers collecting data.

-- Mark

[Evan Balster]

Hello, all —

Regarding what I mean by "computational notebooks" and why I took the time to write a robust math plugin:

I think it's really valuable to be able to interactively model information as part of a note-taking and problem-solving process.  TiddlyWiki was already very good at doing this for qualitative information, so I made effort to round out those capabilities for quantitative information.  My plugin is "spreadsheet-like" because information auto-refreshes like anything else in a TiddlyWiki.  I never actually implemented a spreadsheet UI for it.  Why would I? — tiddlers and fields already work like rows and columns.

I'm a C++ programmer (8 years full-time experience including compiler design, DSP, rendering engines, multithreading), but I don't think imperative languages like C++ or Python are ideal for what I call "computational notetaking".  This is because it takes substantial mental effort to design the execution flow, state management and data structures in an imperative language, and this means breaking away from the topic of the notetaking.  People who use spreadsheets don't think of themselves as programmers because they don't have to do that kind of juggling, but it's possible to design extremely complicated software in that environment.  The key difference is that you can focus on one small element at a time without getting tangled up in the architecture of your program.  Less cognitive load there means more sustained attention toward understanding a problem which may already be straining one's cognitive faculties — and as somebody who gets into that situation often with my research, I think that has the potential to expand my capability to think and solve complex problems.

As much as I've discussed improving TiddlyWiki's scalability, I don't think computational notetaking tools must be useful for designing applications or scalable software.  "Notebook" implies to me that the contents might exist for the sole benefit of the person using it, and that the information is easy to pass around.

(For more on the "interactive modeling" I'm talking about, look into explorable explanations.  I've designed many interactive models for my own use with imperative languages, and have prototyped some explorables with my plugin.)

[TonyM]

Cool Links Evan,

Since before personal computers existed I have dreamed of what computers could be capable of. In the education and learning space I expected to see beautiful representations of complex processes and systems we could play with and thus gain a much deeper understanding. Sadly this is very uncommon and learning has not made much use of computers to this day. Many examples are quite pathetic and not much more than PDF versions :) your examples and work seem to me to be finally moving in the direction I expected as a somewhat naive 11 year old. So sad we still don't have something an 11 year old imagined 30+ years ago.

If someone believes computers have improved the way we learn to a great extent, I would like to suggests it is only because they have a lack of imagination.

Nice Work

Tony

[Sebastian Silva]

Thanks for those wonderful examples! I look forward to doing some similar things as the premise in my project is learning programming by playing!

Today I found the following project which is inspired in Jupyter but uses Javascript:

https://beta.observablehq.com/collection/introduction

It immediately made me think of TiddlyWiki - but I'm not sure how it would fit in. Perhaps as a way to edit tiddlers that is not split-screen.

Regards,

Sebastian

[Ste Wilson]

@Tony.
Agreed.
Let's build something better?
But then it often feels like imagination is punished in education... Certainly in the uk system.. 'what do you mean your not teaching this way? '

[TonyM]

Ste,

I only taught at a vocational college some years ago, part-time. They did not pay us for marking let alone developing material. Why they dont is teaching is a vocation, the capitalist system without guidence, allways under pays and underfunds anything that people are passionate about. They say tgey want the best education, health and science but they simply want the cheapest.

Tony