Yes, they are pure Julia projects. They become regular Julia modules that you can add and run. And yes, full-stack web-apps. Actually, the funny thing is the core web-framework itself is somewhat minimalist; it can do some front-end things, but I designed it to be able to load just about any capability in as an extension. So like core server components are all just extensions. If nothing fits your bill, you can easily just make a new one by sub-typing your own type and doing ?(Toolips.ServerExtension) to find out the consistencies.

Also, it has dynamic routing, server introspection, and as discussed prior does the job of two data formats and one interpreted language. What's better is a full-stack web-app such as the text-editor I built only takes up like 300mb of ram at peak, 150 at idle, and really has nearly no impact on CPU usage.

I'm pretty happy with it. Maybe the source of the project would make things less confusing? A note that I kind of forgot about is that you will need the directory public/texts if you want to be able to actually try the website out.

If you'd like, I could probably even deploy it real quick.

https://github.com/emmettgb/Pasta.jl

There is of course some level of JavaScript that is done on the back-end, there always has to be for the web because the web doesn't ship with a Julia compiler, only Ignition and sometimes other compilers are typically shipped with your web-browser. But the key thing to get is input and output between JavaScript. My design has the document observed by Julia which can then pull out tags as Julia and edit them client-side from there. That being said, any manipulations with those values are performed way faster, and of course you are then connected to the Julia ecosystem as well.

Due to its design, as well, instead of writing a stylesheet, markup, and then writing javascript, the only code involved is Julia -- though you can still include CSS, JS, and HTML code of course.

I don't see why it wouldn't be possible in Python, but frankly Julia made it far easier than Python could have.