It's the "if I can't find it on google or stack overflow it doesn't exist" syndrome

It's not great that it's a common perception. Perhaps "common" is not the right quantifier, but something along those lines.

I've found a lot more people out here in #beginners asking questions. and there are far more #announcements and #news-and-articles these days than there used to be

But it is unlikely to be adopted mainstream

What people neglect to mention is how often ideas from languages like Clojure and Haskell end up affecting the mainstream.

A clear line can be drawn from React's hooks and functional components to reagent and cljs. There are clearly other examples all over.

"PS: Asking because I want to delve into a functional language but I cannot bring myself to do this seriously if it doesn't translate into market value." IMO this is the sign to stop reading; if you let "market value" be the primary constraint on what/how you learn then you'll spin your wheels forever keeping up with the popular framework of $currentyear

it's been said before, but in a labor market defined by supply and demand, "do what everyone else is doing" is not exactly an optimal strategy even for maximizing "market value"

I think the person is just taking the long way to ask if they will be able to get a (Clojure) job if they invest in learning Clojure.

that depends on many other things than just the language itself. places like HN love to overstate the impact of technological choices so they can argue endlessly about their favorite or least favorite language, but it's important to remember that you also invariably have to consider: • the types of jobs you intend to apply for • how you present yourself and your skills in the interview • the relative prevalence of and demand for those skills in your region • etc there's so much else to consider that it's hard to even know where choice of language ranks relative to those other factors.

I would not trust HN to tell me what color the sky is. That is a reality-free zone.

I do sympathize with the poster's perception that the last two years have been a bit tough on FP, but I would point out that there have been some extenuating circumstances (crypto bubble, COVID, funding crunch in VC, economic downturn, managers get more conservative as money gets tighter, etc.), and it was always hard to find actual work in FP.

Every Clojure conference I ever went to, even the sponsors were at best 50/50 on actually doing any. When it came to attendees, the ratio was anywhere from 1:10 to 1:100.

@U03TPPC0R0T what were the sponsors or attendees supposed to be doing or not doing?

What I mean by that is, everyone there was interested in Clojure, but whether they were able to get anyone to pay them to do it, was another matter

The past couple of years excepted, at all the Clojure conferences I've attended over the last decade, there's usually been a show of hands of who's actually using Clojure at work. It used to be just a handful of the audience, but it has grown to be most of the audience over time. So I think the signs are good that "real world" Clojure usage has continued to increase (the conference audiences have also generally gotten bigger and bigger over time too). But if you're not immersed in the Clojure world and not going to conferences, you'd never see that.

From the "outside", Clojure looks extremely niche and many of the public measures of "popularity" indicate a decline (or at best stability) but I think that's just indicative that those measures are not actually very useful for niche technologies 🙂

Gah! So that drew me into the wretched HN site for half an hour, posting "corrections" to several misinformed posts in that thread 😞 This is why I try to ignore HN as much as possible...

orange website bad :large_orange_square:❌

My contrarian take on arguments by reference to relative popularity on google trends- Clojure shows up less because there are far fewer "how to X in Clojure" searches: the standard library allows you to do so much already and devs can just get work done without googling for it ⬆️ that assertion may or may not be true, but it's at least as defensible as any other attempts to judge the health of a language ecosystem by reference to weak signals like google searches

Did you nerd snipe yourself?

defensive coding but for slack

i've the opposite problem, clojure made every other language dead to me.

even babashka is slowly taking over my shell scripts

Measures of language "liveness" from that HN thread and some of the reasons I don't think those measures apply well to Clojure. • Big changes to underlying language ◦ Clojure doesn't need these because most things can be done in libraries • Big changes to big libraries ◦ Most Clojure libraries are small and many Clojure libraries are effectively "complete" and don't need to be changed much. • Endless articles introducing a language and it's benefits. ◦ The reasons that Clojure language was fantastic 10 years ago are still true. The original, articles, books and youtube videos covering Clojure are still valid and Clojure community don't feel the need to repeat itself over and over on these (but maybe we should!).

I find this amusing in the context of complaining that Clojure doesn't change enough: https://www.reddit.com/r/Clojure/comments/xr7n6g/comment/iqdwv05/?utm_source=reddit&utm_medium=web2x&context=3

I am rather sympathetic to "build tooling has become a nightmare i have to relearn every single time I come back to clojure and find all my projects broke without me doing anything"

@U03TPPC0R0T - have you had that experience too? It would be good to know how we can improve this given that tooling stability in Clojure seems quite good to me: (1) The Clojure language is super stable (I've found 10 year old Clojure github projects that have just worked even with new versions of Clojure). (2) Clojars and Maven Central still have lots of old library versions too - so the dependencies aren't usually broken. (3) Some people mistakenly think Leiningen is dead too, but it actually still works great and has significant users for it (the last Clojure survey I saw showed lein had more users than deps.edn ).

I go where the work is first, and the tools second, and the result is my stack feels like it's bounced around like a pinball my entire career. And bitrot is absolutely a thing. I just cloned clojurice to do my semiannual "does this still work" and now somehow brew is installing flipping Xorg. The other day I tried to dust off minicosm for a conversation and found out Figwheel is just broken now, because of course I never upgraded to Main, partly because I never saw a single adequate explanation for why I should, and a lot of "everything broke" reasons not to.

Fair enough - I was thinking of old vanilla Clojure + lein projects that still work years later. But I agree that some more complex builds may take a bit more work to revive them. However, it still feels better than (1) languages with large breaking changes e,g. Python 2-> Python 3 or (2) working but highly verbose tooling like maven in Java.

The thing is, code might not change, Clojure itself might not change ... but the external dependencies might, and Clojure has a very big one: Java itself.

Even the dependency on Java seems quite nicely decoupled. e.g. relatively recent/capable versions of Clojure still work with relatively ancient Java 1.8. In lein you get good control (with exclusions and ordering) and reporting of what dependencies are actually being used. But agree dependency issues do crop up.

A big one that keeps giving me grief right now is "I'm on an M1". So much for "Write once, run everywhere", eh Java? 😄

(And this is a good example of the "I go where the work is, the tools have to follow": I didn't buy an M1, my employer did. I would've preferred not to.)

Fair enough - ARM vs X86 is quite a big change! Colleagues with macs have defo needed to jump back and forth between ARM Java and Rosetta x86 Java - but sometimes even that doesn't work.

I guess it depends on your dependency chain. M1 has mostly worked for me, except one needed hack for RocksDB.

This is very silly, thank you

This is silly. But when you start thinking about how to do it you start to see Clojure's benefits against other languages.

Also there is an important thing about destruction. You kill cells in your body to become stronger. So ecosystem kills languages and people to become stronger as well. And then there is another way to become stronger - - to not kill cells but create stress to them. And then they become stronger individually. So some stress is useful. (thanks, I've read this in a book; these are not my ideas)

The thing I always remember is: > I believe whatever doesn’t kill you, simply makes you stranger. -The Joker, Dark Knight-

I love how @https://www.reddit.com/user/vvwccgz4lh/ is trying really hard to come up with totally serious "solutions" for you. :rolling_on_the_floor_laughing:

@https://www.reddit.com/user/vvwccgz4lh/ is the poster, no?

Yes. And amazingly Slack turned that copy-pasta username into a link to their profile over on Reddit. 🤯

ugh 😄 But well... if we would have enough of these answers then when the "is clojure dead" thread would come up we could simply copy the link to the answer of "let's kill clojure" and call it a day. Also if you're running away from the "let's kill Clojure" kind of thread then Clojure is already dead in your minds 🙂 Because even though everyone says that "Java is dying" or whatever... some people just people don't care. Imagine saying: "we tried killing it, it didn't work" 😄 That would be powerful 😄