Elm's management has left me with the impression of a really smart, well-intentioned hellscape.
Searching for almost any problem I encountered, I'd invariably find two or three Github issues that were exact matches. These were, with impressive consistency, closed about two hours later with a link to some "Meta-Issue for Improvements to our Float handling" or something like it.
That issue, in turn, would usually be closed by some check-in that had nothing to do with my original problem.
Other times, the link to the "Meta-Issue" would get you a 404 because the compiler repository was merged with the documentation repository, and both were subsequently moved to the Elm-Lang project, then renamed to "elmlang". Oh, and feature requests are now part of project elm-future....
Elm is seriously nice to use. But it'd be twice as good if they lost all accounts that have admin privileges for the Github tickets.
A major problem that I see with Elm is that it seems, from the outside, to have a bus factor of 1. Looking at the commits for the compiler and other core libraries you only see one name over and over again: Evans.
Evan has created a very streamlined and pleasurable frontend experience, but IMO his tight control of it will lead to it's demise for a variety of reasons:
1. Something happens to Evan (Hopefully not)
2. Evan decides the foundations need to be replaced (For example if he moves away from the 'Elm Architecture', it will cause a great rewrite and when great rewrites happen developers evaluate other options. e.g. Angular 1 -> Angular 2.)
3. Evan gets burnt out
I'm okay with a BDFL, but not a sole maintainer. I think he needs to raise up others to maintain it for it's long-term health.
Note: Creatively I can understand his desire to control the experience. It also make it feel very unique as a programming language where you are guided in a very narrow path in development, but I believe like most creative works they get better with remixing. When the idea is allowed to be evolved, played with, broken and pushed. Yes, most of those changes will be bad but how else can we know besides trying? Those extensions of Elm's ideas are sadly currently happening outside of the Elm ecosystem because of Evans control, which I think will also lead to it's slow demise.
I think Evan is too protective of the language to let anybody else maintain it right now, the first step would be for him to realize that he should give it up just a bit.
It's a fascinating experiment. A language that's actually a single developer's walled garden, where they make all the decisions, write all the code, and personally vet every library. (There must be other languages where the language team devs hvae to individually approve every package before allowing it to be added to the package manager, but I certainly haven't run across it before!)
It's quite enjoyable to watch, but obviously it's not scalable or sustainable in the general case.
> Elm's management has left me with the impression of a really smart, well-intentioned hellscape.
That seems fair, yes.
My personal view is that there's about a 90% chance the project will collapse under its own weight before it ever hits a release and becomes a normal project, but there's a 10% chance it'll (one day) migrate into a real language (with a real governance model, an actual team working on it, bugfixes being shipped, etc.).
You are wrong. Anyone can publish a package for Elm - without approval.
What you can’t publish is a package with JavaScript in it. And, if you understand Elm, this makes sense.
Elm has no runtime errors - which means any JS has to be perfectly designed. Elm is not to be bound to just a JavaScript platform. Elm has ports for interior.
As for bug fixes - there appear to be some cases where there are bugs, but I have never come across any. The langauge just works.
The slow pace of releases is a business adavatage - we don’t need to be upgrading every N months.
The last upgrade had tools to automate almost the entire process.
Anyway. Would you wager that Elm will fail? I will put up $1000 to your $9000.
Searching for almost any problem I encountered, I'd invariably find two or three Github issues that were exact matches. These were, with impressive consistency, closed about two hours later with a link to some "Meta-Issue for Improvements to our Float handling" or something like it.
That issue, in turn, would usually be closed by some check-in that had nothing to do with my original problem.
Other times, the link to the "Meta-Issue" would get you a 404 because the compiler repository was merged with the documentation repository, and both were subsequently moved to the Elm-Lang project, then renamed to "elmlang". Oh, and feature requests are now part of project elm-future....
Elm is seriously nice to use. But it'd be twice as good if they lost all accounts that have admin privileges for the Github tickets.