5 ms·
I'd encourage Clojure fans to contribute monthly, even in the smallest amount. A healthy ecosystem has a compound effect (like a feedback loop), which can resu
by vemv 9y ago
I'd encourage Clojure fans to contribute monthly, even in the smallest amount.
A healthy ecosystem has a compound effect (like a feedback loop), which can result in a greater abundance of Clojure jobs.
Thankfully Clojure doesn't need to be 'saved' (it's a timeless idea!), but well I often read complaints around its job market and open source ecosystem.
As an example, say you get some VC funding. Can you absolutely trust Clojure(Script) and its tooling? How well will non-senior hires perform? I for one would be tempted to go for something with more traction like Elixir + Reason.
I adore Clojure and all that, but in the end details matter, especially under budget/time constraints.
Of course, this is a (perceived) situation we can change.
- bpicolo 9y ago> Can you absolutely trust Clojure(Script) and its tooling Totally. The closure compiler, despite it's age, is the best Javascript optimizer/uglifier out there. Nothing else comes close. There are several top notch, maintained frameworks as well (re-frame/om)
- vemv 9y agoBut then again, the npm and externs story is not complete yet... Sometimes I get the impression that other transpiled langs have npm + decent minification + decent non-minified compiled size, etc without having to wait for years. Of course, this can be changed (and Clojurists Together can be instrumental)!
- bpicolo 9y agoI don't think the externs story necessarily needs to be complete. The building blocks are all there waiting for you in Clojurescript. You might lose out on a particular library you like, or some random bootstrap + js re-integration, but overall clojurescript has a large number of libraries already available to build top notch frontends. The missing bits are usually small - I need support for <specific analytics library> and those are easy to rewrite, especially given an open-source PoC. The npm ecosystem is absolutely massive, but 99.999% of it is irrelevant for most frontend development. All that said, clojurescript is not necessarily the best choice for any given org. Mostly just saying there's nothing regarding the ecosystem that should hold you back.
- yorwba 9y agoBy closure compiler, do you mean https://github.com/google/closure-compiler https://github.com/google/closure-compiler ? Because I don't think it's related to Clojure in any way.
- snikeris 9y agoClojureScript uses Google's closure compiler.
- fulafel 9y agoThat's the secret sauce that ClojureScript relies upon for tree shaking, dead code elimination and other optimizations. The Closure library is also used for building blocks of the CLJS runtime. https://www.clojurescript.org/about/closure https://www.clojurescript.org/about/closure
- yorwba 9y agoSure, but it can be used on all things JavaScript. I guess I'm just confused why it'd be brought up as specific tooling for ClojureScript.
- KingMob 9y agoThat's actually not true. The Closure compiler has certain restrictions on the Js it accepts for advanced mode compilation, so you can't throw just anything at it. The advantage is best-in-class dead code elimination, which is also why Closure was chosen; to distill the CLojurescript standard library into just the parts needed for a web page requires good DCE. Recently, other Js projects like Rollup and Webpack have gotten good DCE, but when Clojurescript was started, there were no other serious contenders.
- bpicolo 9y agoIn the context of clojurescript you get 'advanced' optimizations free just by writing clojurescript. In JS you have to buy outside the modern JS ecosystem entirely. The closure compiler doesn't play well with any modern JS tooling, but modern JS tooling is nowhere near as capable at dead-code elimination, minification, and code-rewriting.
- dustingetz 9y ago> Can you absolutely trust Clojure(Script) and its tooling? Use Clojure/Script because it's a competitive advantage. If Cognitect went under, my company http://www.hyperfiddle.net/ http://www.hyperfiddle.net/ would still use Clojure/Script and we would, quite frankly, maintain it ourselves, because it really is that important to us. De-risking dependence on any one company is half the point of open source. I'm also interested in the proper way to fund open source, coordinating donations and putting them to effective use is harder than it sounds, it seems to me the easiest way for me to give someone money is for them to sell me something.
- Scarbutt 9y agoCurious, I don't use Elixir/Reason but from a community point of view, how is Elixir more "trusted"? in Clojure you have the java ecosystem which I guess is much bigger than erlang's. Reason still seems to be in high flux?
- vemv 9y ago* Elixir quickly got a bigger job market. Why? It's easier and good-enough, while still benefitial. Also it has a One True framework instead of the Clojure situation (many competitors, different degrees of liveliness) * Sadly JVM-Clojure interop is not used as Rich envisioned, neither technically (we use wrapper libs!) or socially (we create Clojure projects from scratch rather than 'hijack' enterprise Java projects) * Don't know much about Reason. But in general I get the impression other JS communities don't have the same headaches as us (maybe at the cost of some complexity) Clojure is a great bunch of ideas. But ideas also need continuous grunt work and refinement. And finally, IME most programmers will take 'Easy' over 'Simple'. Which is why other new langs seem to have all the fun, and more jobs. Another example: Kotlin. It did in 1 year what Clojure didn't in 10.
- ataturk 9y agoI'd encourage you to just get on with your lives! Before you jump all over me understand that I thoroughly loved writing Clojure back in 2013 timeframe and I did some real projects with it and they were great. It was awesome and what I was born to do. Reality set in soon after. Despite all the hyping by Rich Hickey, Neal Ford and others, the language just hasn't gone anywhere. It is what it is and I'm sorry but it's 2018 and it's done. And Groovy is done. It's just not worth dying on that hill. Maybe go halfway and get some support behind Kotlin if you really want to see jobs available using something besides Vanilla Java 8.
- agumonkey 9y agoLet's trade bitcoins and use the profits for clojure.