1 - active development is relevant. Is it being updated once a year, once a month. Is it going to take 1 year, 2 or 3 years.
2 - Above is not a deal breaker along. But, the new release is going to break stuff. you need to have ability to plan around things, especially as the new release will have paradigm shift (from what I'm gathering). You don't want to write lines and lines of code, then needing to rewrite it all.
I think what makes it even more unreliable, is lack of any communications. I think part of the problem is, Evan (the creator) probably is still not sure what this future looks like. He is very articulate, but not seeing any blog post update makes a reasonable person assume that this new release has not been fully conceived. So, everyone knows there will be a break change. But who knows what will break, when it will be released, etc.
On top of that, this new update has become a gatekeeper for bunch of pull requests and bugs on the existing system that is not being touched.
The people who sit next to Evan the author of Elm in Action and the biggest users of Elm (Richard Feldman) has pushed back the release of his book by one year. Basically the remaining part of his book, is all the SPA stuff that the version of Elm is supposed deal with. From what I've seen, it seems like even he doesn't know what to expect [0]
Put everything I mentioned together, it makes it hard to recommend Elm for anything your company will be depending on, especially when there are other solutions out there. Why take the risk.
Every single version of Elm has had breaking changes. However, 0.17 -> 0.18 had an update tool that automatically fixed most of those changes, and a tool for that will be available for 0.19 as well.
NoRedInk (where Feldman works) are already updating a bunch of code to 0.19, according to Luke's talk at Elm Europe 2018, so they seem pretty sure of where things are going. Evan doesn't like announcing things until they are done though.
Because Elm is still a young language that hasn't found a solid footing that operates as a core philosophy, and usability issues and other concerns go unanswered or ignored for years at a time. When a language is very opinionated, and those opinions change, it creates a stability issue, and active releases and engagement with users help mitigate loss of confidence.
It matters for planning. If I knew when the new release is coming, and which features it brings, then I can decide whether I use the current stable version, and upgrade later if required, or just use the beta version during the development knowing it will go live in time for my go live.
Knowing just that there will be some changes, but having no information about he final scope or timeframe makes the decision much harder to take, and can just lead to leaving to the competing product / language / library / framework with more predictable timelines.
Good point. Spare a thought for anyone who's gone through the various releases of Angular in the last 18 months. The other points may be valid though...
1 - active development is relevant. Is it being updated once a year, once a month. Is it going to take 1 year, 2 or 3 years.
2 - Above is not a deal breaker along. But, the new release is going to break stuff. you need to have ability to plan around things, especially as the new release will have paradigm shift (from what I'm gathering). You don't want to write lines and lines of code, then needing to rewrite it all.
I think what makes it even more unreliable, is lack of any communications. I think part of the problem is, Evan (the creator) probably is still not sure what this future looks like. He is very articulate, but not seeing any blog post update makes a reasonable person assume that this new release has not been fully conceived. So, everyone knows there will be a break change. But who knows what will break, when it will be released, etc.
On top of that, this new update has become a gatekeeper for bunch of pull requests and bugs on the existing system that is not being touched.
The people who sit next to Evan the author of Elm in Action and the biggest users of Elm (Richard Feldman) has pushed back the release of his book by one year. Basically the remaining part of his book, is all the SPA stuff that the version of Elm is supposed deal with. From what I've seen, it seems like even he doesn't know what to expect [0]
Put everything I mentioned together, it makes it hard to recommend Elm for anything your company will be depending on, especially when there are other solutions out there. Why take the risk.
[0] https://www.manning.com/books/elm-in-action