Of course. I’m not sure why it’s so hard to get my point across.
By the time your startup idea proves to be a failure or your tech decisions turn out to be disastrous, you might already have burnt a couple of millions.
“I get paid” is not a metric of even partial success of the company in any way.
Aha, so you’re saying “any company that manages to pay an employee is successful”? In that case that’s probably 99% of all companies, regardless of whether they go broke, waste investor money or are just terribly mismanaged despite surviving.
Not very useful.
You are saying
Do you mean by most 99%, 90% or even 75% ?
Well in case, that statement is incorrect. Most companies which have a record of surviving (I’m not talking about startup) make profit . Statistically a company which is alive as more chance to be making profit than the opposite.
You seem have little to no experience in startups, startups in fact do pay a pittance or don’t pay at all, they provide stock and options to make up for it but realistically 99% of these go to near 0, aren’t liquid and are subject to some cliff periods.
Success isn’t a one off thing, it’s continuous, as is everything in life.
This particular line of discussion is becoming unproductive very quickly. I suggest we don’t prolong it more than necessary.
I can agree that no tool/language can suit every use-case; but, to me, “it isn’t for everybody” comes across as gate-keeping.
I want Haskell to be fore everybody. I do belive the world of software development would be better if everyone knew and/or used Haskell. Whether this requires changing Haskell or changing the world, I am up for the challenge 
Gatekeeping is intentionally making something inaccessible for the purpose of keeping people out.
That doesn’t fit what I’m saying about Haskell. It really isn’t for everyone and everything. Some people and companies have values that clash with the language, for instance. And that’s fine.
I didn’t assume (or mean to imply) this was what you indented to say. I only described how it came across to me.
I worked 1.5 years in a Haskell startup, which later failed due to business reasons. I was also the head of engineering for a while. Feel free to look up my CV.
At no point in time did any engineer work for free.
Please refrain from ad hominem arguments.
When it comes to people, Haskell is actually suitable for everyone (see https://www.haskellforall.com ), just not necessarily good / idiomatic Haskell. The ability to do computation in IO, however rancid and smelly, means that Haskell can be reduced to Fortran if the user absolutely demands it, but we hope that dyed-in-wool imperative programmers at least learn proper effect / computation separation during their stay in Haskell-land.
As for companies, values is very moralistic. Companies have different requirements; Hasura, for instance, seemed to have wanted to play with the hot new thing (Rust), wanted the performance improvements, and didn’t want to deal with Haskell’s space leak issues anymore.
Some firms might want ultra-fast prototyping, and for that, a Lisp like Clojure might be a good solution (dynamic typing and homoiconic macros, we are taking bets on how long it’ll take for the Clojurian in the thread to hit a complexity wall).
Is this a quote from someone at Hasura or just speculation?
Because space leaks are a “once every couple years” fuckup and not something to rejigger your entire engineering corps over 
I feel this thread has very interesting posts (Chris’ et al.) from lived experience. Alas, there is more and more pointless bickering in more recent messages, so I am going to lock this.
If you feel you have anything relevant — and polite! — to add, feel free to start a new discussion.