I’ve just began a new role in a startup organization, we are building an application centered around a UI that needs to be heavily integrated with Cesium.js.
As of yet we have tentatively decided to use React. As you may have guessed, I’m really pushing for Elm. With in six months we need a working application that can not break during demonstrations.
There are two road blocks to using Elm.
FFI may make integration with Cesium more difficult
Developer resources
I would like any suggestions of how to make the Cesium integration more streamlined, however I mainly would like to know:
How big is the Elm community in Victoria, Australia?
What is the expected salary? (ball park)
If you are interested, let me know below! I need to make a case for Elm.
Hi. Looking at Cesium.js I feel like you’d use custom elements for communication. In any case FFI with JS is super simple in Elm. That’s should be a non stopper for sure.
How big is the Elm community in Victoria, Australia?
Rather than answer “How many potential applicants know X?”, I’d turn it around, ask “What does knowing X tell us about an applicant?”
When X is not yet mainstream, it’s a little extra signal to help identify those that are motivated or curious enough to learn for themselves; picking good applicants is hard enough without it.
It doesn’t take long to bring others up to speed, either; we’ve had devs learn & become productive on a >160kloc Elm app within their first week.
(Also working fulltime on Elm in Sydney, but that’s not quite Vic, eh? )
You might want to take a look at my package gampleman/elm-mapbox, which shows a fairly full featured integration with MapboxGL (which is also a 3D capable mapping library as you probably know).
What I can say from building mapping apps in React and Elm:
Elm was way more awesome and maintainable in the end. No surprises here.
The React wrappers we found for Mapbox were always quite awkward, as Mapbox used a bit of a different mental model than React. I have no idea if this applies to Cesium.
The wrapper library for Elm I ended up writing made #1 above possible, but was a fairly significant time investment (in this case ameliorated by Mapbox providing a decent chunk of their API as a simple JSON schema, so I used code generation heavily).
As others have suggested, the main thing I would worry about is if other developers in the organisation are happy to learn Elm. You can likely hire (especially if you don’t actually require candidates to already know Elm, but rather be proficient in frontend), but you want to avoid a fight with the people already in the org, as that will slow you down.
I used to run the Melbourne Elm meetup (last meetup was Feb 2020 because Covid).
We usually got around 8-15 people every month, lots of new faces each month interested in Elm.
From the meetup I can only name two companies actively developing in Elm in Melbourne (CultureAmp and Stax).
My guess is something between 130-170k AUD per year.
)