Original title -- "Ranking Programming Languages by GitHub Users" -- is much less clickbaity that the current one, which seems to have an agenda behind it (at this moment is "All mainstream languages marketed as functional are losing market")
I didn’t write that with an agenda, I wanted to point out this data analysis which is a common topic for the HN community; unfortunately I could only link to the top the page and not just to the section.
The title of this post (on HN, not the original article title) leaves out the rest of author's sentiment on the data.
"Given the relatively small numbers here, there is more noise in the rankings. I'm also not convinced that the apparent decline of things like Clojure and Haskell is anything more than a result of GitHub growing more mainstream over time."
This title (for posterity: "All mainstream languages marketed as functional are losing market") is extremely editorialized and is even contradictory to the text of the article. Please don't do this.
Unfortunately I could not link to the section which actually shows this data. I though the topic of FP and market share is relevant for the HN community where FP is often a big topic... the rest of the article is programming languages without any especial surprise.
P.S. can some admin edit the title and add “(see the last section)” so the topic of less users of functional languages can be spoken and explored in the community?
I think you tried to steer the discussion too much instead of justing let people read the text and reach their own conclusions. Which might be nothing like yours.
Now, as a result, the post is flagged and there is no discussion at all.
In the future, I suggest submitting the article with a non-editorialized title, as the rules stipulate, and then starting a comment thread to spur discussion.
The graphs for major programming languages shows a decline in the number of monthly active users using JavaScript, interestingly this decline seemingly coincides with the incline of TypeScript users. I'd love to see whether people who stopped using JavaScript started using TypeScript instead.
Of course! When my company decided to embrace javascript, I spoke out against javascript and pushed for typescript, the entire engineering team is writing typescript instead of javascript.
It's also a bit strange in that Flow and TypeScript are competing JS language extensions with nearly the same goals and scope, yet TypeScript is seen as its own language and Flow isn't. And, of course, JSX isn't valid JavaScript, but React programmers view themselves as writing code in JS. IMO, it would be better to put TypeScript projects in the JavaScript category and optionally have more detailed stats on the different JavaScript variants.
When I introduced TypeScript to my team at work, one of the things that I had to emphasize was that TypeScript isn't a new language; it looks and behaves just like JS, it just has some new syntax to declare types.
I suppose the practical matter here is that GitHub language stats can distinguish TypeScript because it uses the .ts extension, while other JS variants just stick with .js (and JSX sometimes is .jsx, but not always).
This is data, not even a model, and interpretation is hard. You may see the decline - of e.g. PHP - as an opportunity or a threat. Anecdote: in Finland there is one person providing APL training. Guess how much insurance companies pay him ;)
I guess the accuracy of this characterization rests on what exactly is meant by "mainstream languages marketed as functional" - specifically, the "marketed" bit - but I would argue that the mainstream trend is to embrace functional programming. Python, Kotlin, and Golang all support functional programming in non-trivial ways. Java streams, lambdas, and immutables each represent huge shifts towards functional programming style. C++11 and C# 3.0 have added similar constructs (according to Wikipedia, at least - I am less familiar with these languages). So while the relative interest in 'academically pure' functional languages may be declining, it seems to me that stylistically functional programming is becoming more popular than ever.
I'd agree with you regarding Python and Kotlin, but Go is pretty clearly anti-functional. It does include closures but little else, and Rob Pike is pretty clear as to the preference for imperative (e.g. for loops) over functional (e.g. map/reduce/etc.) style.
Fair enough, I don't use Go. I'll admit I just googled "Golang functional", skimmed some stuff, and assumed it fell into the same bucket as Python & Kotlin.
I would say those ones are not marketed as functional, but they do support functional constructs; same as JavaScript. Elixir for example does not market itself as functional but as a niche language for distributed systems which happens to be functional.
Elixir does not market itself just as a functional langauge but as a niche language for distributed systems which happens to be functional for very good reasons. As opposite clojure markets itself as functional in spite of having many elements of imperative languages and sitting on top of the JVM.
21 comments
[ 3.3 ms ] story [ 18.6 ms ] threadActually I use FP languages in my free time.
"Given the relatively small numbers here, there is more noise in the rankings. I'm also not convinced that the apparent decline of things like Clojure and Haskell is anything more than a result of GitHub growing more mainstream over time."
P.S. can some admin edit the title and add “(see the last section)” so the topic of less users of functional languages can be spoken and explored in the community?
Now, as a result, the post is flagged and there is no discussion at all.
When I introduced TypeScript to my team at work, one of the things that I had to emphasize was that TypeScript isn't a new language; it looks and behaves just like JS, it just has some new syntax to declare types.
I suppose the practical matter here is that GitHub language stats can distinguish TypeScript because it uses the .ts extension, while other JS variants just stick with .js (and JSX sometimes is .jsx, but not always).