And also, Miso is quite weird, because those functions are only exported in the GHCJS version of it, not on the regular GHC one
So we would have to separate completely the backend from the frontend
Shoudlnt be that hard tho
ha, what I suspected
now I'm trying to write a bare-bones project to figure out what goes where
I do like miso because M-V-C are well separated
though I'm completely ignorant re. frontend development
I feel miso is a very good choice to refurbish the project
Indeed it is @ocramz . Another thing that I was thinking of was to completely rewrite the backend and make it a RESTful API using some other framework like wai or yesod which are much more used in the Haskell community
@NickSeagull what functionality would you put in the backend? the filesystem browser?
I'm having a good experience with miso but I don't get why certain IO doesn't work, for example listing a directory
so I wonder if it's a good idea to move that to a pure GHC backend
once I figure this out I'm ready to help with the haskell-do refactor
I was trying to figure out what miso is without digging into the code. What is it supposed to do?
@dataopt it lets you write both the visible part and the behaviour of a single-page application
in the "model-view-controller" style, i.e. a finite state machine
the model is the state, buttons and whatever inputs are the controller and view is the visible part of this
@ocramz I would basically put the filesystem browser, compilation, save file, etc...
@ocramz would be interested in what you might have in mind.
I've been thinking to put a websocket in chart-unit for a while now, then luna came along, and now I want to try something like that out.
I imagine it's beyond svg, so would have to learn about front-end and canvas, and miso might wrap it all up nicely.
@tonyday567 you might check the transient gitter for some of @agocorona 's latest notes on that.