I just saw this on slashdot.org
Here is the story:
Hell, MIT julia lab is so big?!
Okay, let’s have a conversation about this then. What does everyone here want to see running in the browser?
I would personally like to see Plots/Makie and WebIO working really nicely on the client side so that I can build a web application for playing around with Flux and MLJ models, including graphically manipulating them when I’m not near a desktop/laptop computer but could use a tablet to do simple manipulations to my models (with the option to run the model in the browser, although I care about this less).
I wonder what’s the the relation between this and a jupyter wiget
I want a way to let people try Julia in the browser that scales trivially and doesn’t cost too much money to offer for free.
Me too ;). Plus all the cool reactive GUIs that people built but didn’t make sense over the web because of the large amount of communication between them might suddenly start being doable even on the web.
Would be nice to play games on the browser built with Julia.
So, FF would be the second (but best) web browser running julia. I thought this would be a great place to point to my toy project.
It’s not really ready to show off but since the subject came up…
I’ve been following iodide with great interest so it’s really awesome to see this. Congratulations @vchuravy !
In my ideal world it would be great if we could have a way to distribute interactive but pre-canned data visualizations which can run in a browser without server-side support. For example to report statistical analyses to non-technical stakeholders without maintaining server side infrastructure. Iodide seems quite close to what I’d want to use.
It would also be nice to lower the barrier to entry for writing fast GUI applications in the browser by allowing the client side logic to be written in Julia. The complexity of the modern client side stack seems quite excessive compared to whipping up a simple GUI in Gtk.jl.
I fear that, for it to happen, we first need a standardised way to fully compile Julia code into executables, entirely removing any JIT overhead.