The idea I'm having is that by using JS+HTML+CSS+<BACKEND{JS, Python, Go, ...}>[+SQL] instead of just Elm I'm prioritizing my complexity fetish over real-world cost/benefit breakdown.
I don't like Elm in the same way I like Python, for example, but I cannot make a case to myself for NOT using it. Anybody got reasons?
I find this very interesting. There are lots of facets here:
1. Is anyone else ever going to work on this code base? Probably so. You have to factor in the Elm learning curve (since they are unlikely to know Elm).
2. Elm abstracts a lot of underlying web technologies away from your control. Does it work well? What do you lose in terms of control, speed, or ability to debug?
3. What are you comparing Elm to? It doesn't seem like it to us HN users, but Python is almost a niche language for the web (compared to C#, Java, JavaScript, and PHP). It's harder to argue for Python as a result, but what about a more popular language?
4. All of this might be moot because I've heard Elm's "time-traveling" debugger is absolutely amazing: http://debug.elm-lang.org/