Hacker News new | past | comments | ask | show | jobs | submit login

Eh, it doesn't seem like character assassination. Dude chose his management style. That makes his project unfit for purpose for certain people. Those people are allowed to make that determination and state as such.

I remember when the Elm team first decided to lock down the Elm compiler, so that only the team in charge had special privileged access to modify it. I thought, "wow, this looks like a bunch of people who don't want to collaborate, who want to take their ball and tell others how to play with it, really putting the 'dictator' in BDFL." and you know what? that's fine. I don't harbor anything against them as people. But that left a permanently bad taste in my mouth and I'm unlikely to ever trust that group or a project they're managing.




Without this management style Elm would be the oppositive of what it is today (simple language, very fast compiler, efficient tree shaking, no runtime exceptions) wanting to add at all costs Haskell features and JS direct sync interop.


Because of that management style, it has been bitrotting for FOUR YEARS and has even gotten forked by a couple people involved so at least some progress could continue.


I simply don't think locking down the compiler so only blessed Elm devs can modify internals was necessary. I don't think the flack given to people forking the repo was necessary. Say no to pull requests, fine. Don't try to control the technology.


From outside, with no knowledge of this "ELM Controversy", this is all very confusing.

I love the ELM Architecture. Why is there a problem if it is implemented in another language?

I'd think any functional language could implement the ELM Architecture, and thus be 'native', or more integrated into that particular languages ecosystem, and this would offer benefits beyond trying to integrate ELM into each individual ecosystem.

And doing this re-implementation, would in no way be a comment on the ELM founder.

Also, Doesn't Linus say "NO", quite a bit with Linux? Why jump on the ELM guy for doing same thing.


When Linus says no, he rejects features and pull requests. That's fine. That's responsibly shepherding your implementation.

That's not what Elm does. They lock down the compiler so only Elm devs can modify internals, which would be equivalent to Linus trying to prevent people from modifying the kernel.

They really don't like people forking their repo, which would be equivalent to Linus getting mad at people making derivative kernels (because it would "split the community.")

It's not just a matter of saying "no," it's trying to control what devs do to an UNREASONABLE degree.


The link is that Elmish is a competitor of Elm, and that the creator of Elm is a factor to the evolution of Elm. Leadership is a big deal in open source and is the sole differentiator behind the late iojs and OpenTofu. The matter of leadership is especially weighty for niche or emerging ecosystems like Clojure, Rust, or SolidJS.


FWIW my comment was along the lines of allergic reaction. I'm not an Elm fanatic, I'm just sick of seeing people snipe on the dude.

You're entitled to your opinion, of course, and you articulated it well. But again I see it as another example of someone not liking being told "no", no?


That's really reductive and biased against us, the potential user base of Elm that decided not to engage. Why is it that we don't like being told "no," as opposed to the Elm leadership and Elm community not liking being told "this isn't a 100% solution for me so I'd like to modify it"?

Obviously Elm is his passion project, that's fine. But nobody who locks down their compiler on a supposedly "open source" project can actually seriously expect others to engage with that in good faith.

Furthermore, the absolute state of just trying to modify Elm to be more fit for purpose is horrendous. Every time I see a parallel implementation or a fork, the author has to bend over backward with platitudes to placate the Elm masses and ensure them that this project isn't trying to "split the community," which is a nonsense idea to start with imo.

It feels cultish, on top of a project that is actively hostile to people who want to use it any way other than how the BDFL has mandated its usage. If you want to say the problem is with the users, that's fine, but that's exactly the attitude keeping me from ever engaging. I personally think the problem is with the people who want such fine control over their technology and community, as opposed to understanding that they can really only shepherd their implementation. That's about as far as their reach rightfully extends.

EDIT: also, the fact that they aggressively lock down their compiler and have such hostility to anyone working in parallel is imo the definition of "holding it back"


> That's really reductive and biased against us, the potential user base of Elm that decided not to engage.

My complaint isn't that you or anyone doesn't engage with Elm, that's your prerogative, of course.

My problem is that a small segment of people do engage with it whenever it comes up to smear the project in general or Evan in particular for nothing worse than wanting to run his own project his own way.

He's not an ogre? I mean here he is, look, he semms nice enough to me? "The Economics of Programming Languages" by Evan Czaplicki (Strange Loop 2023) https://www.youtube.com/watch?v=XZ3w_jec1v8


As far as I've seen, there was no attack on the person.

What I usually see are warnings about what happened to the project and how it is run, mostly targeted to the unaware. This is a sign of respect to people's time before they waste it on something that might disappoint them when they get to know the details.


Every time Elm comes up people climb out of the woodwork to cast aspersions on it for no better reason than that Evan wants to run his project his way. Some people evidently just can't accept that and have sustained a grudge that drives them to comment on it whenever it comes up.

Look, I get it, I'm still sore about Google Feed Reader, 'member that?

But enough is enough. It's old news. Get over it. Etc.

It's like if every time someone mentioned Caddy server there was someone bringing up that time Matt Holt put that header in and pissed off a bunch of people. Life goes on. We forgive and let things go.

It's high time for folks to forgive Evan for his mistakes. I mean, he's very young still, eh? Cut the guy some slack already?


Again, the things I read are not personal, grudges or inability to accept someone's personal choices. For starters, you write his name more than anyone else here. You're not doing him any favor specially when it comes to Streisand Effect.

When I propose a tool to my CTO I couldn't care less about the project's leadership personal choices.

Mature engineers think in terms of ROI, dependability, bus factor, predictability, funding, mind-share, hiring ease, license, etc.

But it would be naive and unprofessional of me to handwave technical issues with the tools. And ALL tools have hard edges and undesirable characteristics.

And putting a header in a HTTP server is not proportional as restricting who can change the source code of a compiler.

Matt Holt is awesome btw. And so is Elm author.


Evan wants to control the technology. That's a management decision that makes it unfit for purpose. This is a legitimate criticism that had long ranging impacts. I have seen no reason, no commitment from the management side, to believe this won't happen again. It is worth bringing up when there are similar alternatives without such a spotted history of lockdown and control issues from management. I know nothing about the fellow. This is not personal.


My understanding of this is that he managed to hype the language with certain killer features. A lot of people were enthusiastic and invested alot of time in learning and building in it. Evan than "suddenly" said they wanted to take the language in another direction and dropped some of the core features and pricipals that drew people in. While perhaps a mischaracterization, it comes a cross as a kind of "bait and switch" or at least undependable.




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: