sciter-js-sdk VS neuronika

How tiny is the output? What is the API? (I couldn't find any documentation. I thought maybe it's in the Help menu in the app itself but all I found was this https://i.imgur.com/6puOMIe.png - On that note, what level of JS is supported? As far as I can tell it uses this https://github.com/c-smile/sciter-js-sdk which is based on Fabrice Bellard's QuickJS ) Is there sound?

Having normal procedural language for build automation is of course useful. That's for those 10% of cases when "standard" build DSL (make,CMake,etc.) simply do not have facilities.

But the rest of 90% tasks should have compact (easily readable) definitions.

I personally found that Premake5 has quite good balance for these tasks.

Premake files are plain .lua files and due to Lua syntax they are easily readable. And if needed you can call from them procedures defined in again Lua.

So typical project (multiplatform) definition looks pretty readable, for example one project from Sciter SDK:

https://github.com/c-smile/sciter-js-sdk/blob/main/premake5....

More details.

That's why in Sciter I've extended JSX with translation meta instruction @:

Note Tauri is full fledged Client/Server with WebView (client) is running in separate process with RPC between UI process and Rust code (Server).

For the comparison:

Standalone Sciter (scapp.exe, https://github.com/c-smile/sciter-js-sdk/tree/main/bin) takes ~8 MB of RAM (with minimal Cairo and GDI backends).

That's 20 times less than even Tauri.

WebView based solutions are not suitable for applets - small portable desktop applications.

scapp.exe ( Standalone sciter engine ) takes 45 Mb showing its default "about" document.

Check this example of native object: https://github.com/c-smile/sciter-js-sdk/blob/main/demos/int...

So you really don't need V8 JIT infrastructure to achieve max performance - just add native objects when needed.

Also perhaps comparing to Neuronika.

how does it compare with https://github.com/spearow/juice, https://github.com/neuronika/neuronika and https://github.com/spearow/juice?

Check where and how this struct is used. https://github.com/neuronika/neuronika/blob/variable-rework/neuronika-variable/src/history.rs

I have a question: as the maintainer of [neuronika](https://github.com/neuronika/neuronika), a crate that offers dynamic neural network and auto-differentiation with dynamic graphs, I'm looking at a future possible feature for such framework consisting in the possibility of compiling models, getting thus rid of the "dynamic" part, which is not always needed. This would speed the inference and training times quite a bit.

If you like autograd you will love neuronika

After reading most of the book, I wanted to get my hands dirty. My initial idea was to build a small machine learning framework but I deemed it to be too difficult if not impossible for me at the time. (Now, neuronika would have something to say). When gathering the bibliography for my thesis, I recalled to have stumbled upon a particular algorithm, t-SNE, whom I liked very much. I found the idea behind it to be very clever and elegant (t-SNE it's still one of my favorite algorithms, together with backprop and SOM, I find manifold learning fascinating in general). "So be it", I said, and I began writing a mess of a code, that was basically a translation of the C++ implementation. Boy was it bad.

me and a colleague of mine are developing neuronika