Why Clojure?(blog.cleancoder.com) |
Why Clojure?(blog.cleancoder.com) |
I guess he's happy. At least he isn't touting Haskell. Bon Voyage, Bob!
But somebody needs to break it to him that Lisp is not a functional language. Or, if it is, so is C++.
You're missing the point. Everybody can write Fortran in any language.
Lisp is not a language at all. It's an idea. And if you grok the awesomeness of that idea you can make any kind of language using it: functional, OOP, relational, whatever.
Bob is talking about Clojure. Clojure is a functional Lisp.
I also disagree with you on C++. Cc+11 was a fantastic improvement, C++14 added some missing stuff — at this point I was sold and eagerly awaiting C++17 — but when 17 arrivd and when I saw what 20 adds... it has become so complex that I don’t understand it at all anymore. I used to scoff at the people who said C++ was too complex, but after ~18 years of using the language its now reached a point where I am no longer comfortable using it. I see code that uses C++17 and recently 20 and I have no idea what its doing anymore. I am no longer confident that I can write safe, non bug-ridden code in C++.
Did you have prior experience with dynamic languages? If so, how much?
Did you have prior experience with statically typed languages? If so, how much?
I'm trying to see if Clojure requires more fundamental programming intuition and experience to sustain in large projects. I work with tens of thousands of Clojure LOC, I don't feel these issues and can't relate, and love Clojure, so I'm curious to understand what context it best applies too. I wouldn't want to force it on a team that wouldn't benefit from it, so I'm interested about learning these aspects, so I'm able to recognize in what context it would make sense for me to influence a team to adopt it or not.
Thanks.
In Erlang/Elixir, you have patterns, and interfaces which are fairly low level, but provide meaningful abstractions over common goals... I.e. we have OTP. OTP is fairly simple, and that doesn't mean it's easy, it's really not, but it can be simple once you're familiar with it.
Rust, on the other hand, is easy to get started with, the core of the language is fairly small (and relatively simple) but that's where the complexity starts to creep in.
It's so small, I have to constantly reach for a crate for common goals... There are tons, and they're fucking awesome, BUT they're easy, not simple, and my application's complexity grows exponentially relative to my familiarity each time I grab another crate.
I think it's easy ;) to say the Rust ecosystem has to contain the complexity it does, and making it easy is the best we can do, but with great challenges comes great reward... Batteries not included, yes, but at least you could know the size (spec) required for the ones you need? That's probably a shitty metaphor but I tried...
...
One more thought on this tirade, when Rich Hickey casually says that engineers should be working to remove complexity from the business... That hit me right in the feels. Too often do we get complex business decisions/goals and just accept them. We need to push back and help them simplify their goals... too often I've gotten some crazy request, where instead of implementing it, I pushed back to get to the root of the problem and ended up with something wildly simpler and better for both parties. When we push back, it shouldn't be to make something easy, it should be to make something simple. You can scale simplicity almost infinitely, but complexity will eventually come crashing down.
More on topic... Thanks for posting this talk, it and the article have inspired me, I'm gonna go try to write a bit of Clojure today. You should write some Elixir, or maybe LFE (Lisp Flavored Erlang), I think you'd quite like BEAM :)
At my last job the client reps were notorious for saying “yes” to every client request and other devs were notorious for implementing those requests without question. I really tried hard to push back and find out what the client was trying to achieve. So many times after a brief discussion I would be able to inform them that what they wanted was already possible or doable with way less effort and expense. Sure, we didn’t get to bill them as much, but I’d much rather solve their pain points without adding code (especially when the code has already been written) than needlessly charging them.
Would love to hear why? What is that make Clojure such a good experience for you?
What's made clojure so great, imo, is its unicorn status as a principled-yet-practical language. That "principled" part is not worthless---it means that a lot of great minds are drawn to it. Before react took over the world, clojure folks were already taking steps in that direction.
A lot of other things. The "sequence" as a core abstraction is very powerful. Immutable data structures by default make functional programming perfomant and efficient.
Speaking of data structures, data structure literals have spoiled me for other languages. After Java, especially.
And none of this mentions clojure's lispiness. The ease of metaprogramming has allowed the community to build some of the best tooling out there, between CIDER for emacs, and Figwheel for the browser---oh, did I forget to mention clojurescript? Being able to reuse code on the front and backend is great for web applications.
Clojure isn't "everything." I think I'd still benefit from learning Haskell, APL, and Forth, and I wouldn't mind knowing Ruby and js a bit better. And I'll probably be dragged back to Python and R if I keep doing maths.
But if someone asked me which single language would probably do the most for them professionally, I'd say clojure. It's the language people start startups so they can use it.
- Immutable data-structures with concise literals for lists, vectors, maps, and sets. Having pure functions and immutable data-structures makes code easier to reason about, easier to test, and thread-safe.
(But clojure doesn't "force" you to be pure. The idea is that you write as much of your code in pure functions as you can, and push the IO and impure parts to the extremities. It's very pragmatic in this way, and you can get real work done.)
- Simple syntax:
While the syntax may be intimidating at first and arithmetic looks weird to untrained eye, once you realize that everything is a function (or a special form that also looks just like a function), the very light syntax and consistency feels amazing/refreshing.
- Macro system:
This is tied to the previous point. Since all your code is technically a "list", and you have something akin to a pre-processor where you have the full library of clojure functions to manipulate data. But in this pre-processing phase, your data is the code. (the code is a list). Now you can dynamically re-arrange or re-write code. Lookup homoiconicty and learn about the kinds of things you can do in macros that can't be done in other languages.
- Capabilitis for general programming, abstractions, code re-use etc.
- Performance is quite good. It's often nearly as fast as java, yet your codebase might be 10x smaller because the language is so expressive.
- Concise/expressiveness. Clojure comes with a nice built-in library with generic functions that you re-use again and again. Give a programmer a few dozen of these functions and it's amazing what can be composed to solve many problems succinctly and elegantly.
- There are surely other benefits, but the last one I'll leave with is hard to explain unless you have felt it before. It's REPL driven development.
Clojure comes with some seriously awesome REPLs (i.e. a shell for interactive tinkering with the language). Other languages may have some form of REPL, but no other language in my experience has come close to the feel you get in a clojure REPL. I can best explain it as a freedom of very light-weight experimentation that you use to write your code. It's a playground for writing functions with very quick feedback to see if your code will work or not. One factor that makes the clojure repl experience so nice ties back again to the succinctness and expressiveness of the language. Typing commands in the repl is painless because it's concise, not a lot to type, and then the feedback is so instant.
Somehow when I write code in python, java, javascript, or other languages, I just don't use their REPL or shell as often. It's just not quite the same as the clojure repl experience.
Then, when I learned about Elixir's macro system, I thought, "wait a minute, this is LISP! This is just a LISP in Ruby's clothing in Erlang's pasture!"
That was a delight. :) My heart still belongs to the LISP family, but right now I'm working as an Elixir developer.
One of the general arguments in Lisp vs the world is that lisp is more concise. F# is really concise. It has datastructures as intrinsic part of the language syntax (just like Clojure, i.e in F# [|,,,|] is an array and in Clojure [...] is a vector, so more or less the same thing). Furthermore, the type inferences concierges the developer through the regular bureaucracy involved with strong type systems so effortlessly that you seldom notice it ... unless your code does not compile, in which case it's wrong somewhere and you need to fix it (which you needed to do anyway).
Ok, so for the sake of the argument, let's say the syntax is equally light for F# and Clojure, and their support for immutability is in the same ballpark. And let's remove the argument "a type system is an encumbrance" because that will only lead to situation specific debates - one side will say that massive refactorings are a pita without types, and the other side responds that in Clojure you can use data schemas and should have unit tests anyway. And those discussions never lead anywhere conclusive.
So, besides the above mentioned topics, is there anything general one can say of the distinctive advantages of these two languages in specific situations, including how their two different ecosystems and runtimes affect the calculus?
I suppose it's extremely rare to have experts who have participated in any large scale industrial developments with both of the languages, but if there are, it would be really nice to hear their opinion. And the things I listed as "outside of the discussion" might actually be critical, who knows...
> No. Clojure is not slow. Oh, look, it’s not C. It’s not assembler. If nanoseconds are your concern than you probably don’t want Clojure in your innermost loops. You also probably don’t want Java, or C#. But 99.9% of the software we write nowadays has no need of nanosecond performance. I’ve built a real time, GUI based, animated space war game using Clojure. I could keep the frame rates up in the high 20s even with hundreds of objects on the screen. Clojure is not slow.
If you're going to respond to this question at least provide some comparison maybe.
But otherwise that anecdote of "I could keep the frame rates up in the high 20s even with hundreds of objects on the screen." absolutely screams extremely slow. High 20s FPS with hundreds of objects is incredibly bad on the surface of things. Maybe it's closures fault, maybe it isn't and is instead the fault of whatever was doing the drawing. But it's not a good look either way if that's the only data point that can be provided.
Similarly the IDE/dynamic type question seems to have missed the main downside of dynamic languages - lack of good IDE & other tooling support. How well are things auto-completed for me? How well does the IDE warn I've made a type error before I've gone through a long compile & test iteration loop? How well do linters or other things work, or is there some aspect of LISP that means this is just not an issue like it is in other languages?
Funny, this list is the same one I use as to why I'm so annoyed with Clojure right now.
I inherited a mission-critical Clojure ML library my team uses for it's primary business goals. It was written 4 years ago by a research scientist- who quit 3 years ago. We know what it's supposed to do. We know that it seems to do the job well. We just can't understand the code well enough to be certain of what it's doing or how it's doing it. If this thing breaks or stops working, we're screwed.
The problem is that the author, like most of us, found great joy in writing very few characters to express very big ideas. Clojure let him do that to an extreme degree. And I'm sure if you were sitting beside the author, with him explaining these dense expressions, you would be enlightened at the elegance and beauty of this language.
Sadly, I was not.
I'm sure Clojure is a lovely language. But it lets you write a Voynich manuscript that compiles.
For me, Clojure and JavaScript (at least for modern React community) are similar when it comes to usage: Dynamic typing with Hash-map centric modeling, use functions to compose thing but mostly without monadic patterns. The methodology in the Clojure community has an official name: "It's just data". This methodology works, and have it's own pros and cons. The obvious pro is people can write generic functions without poking hundreds of nominal types, interfaces and type classes.
The only big difference I see is macro, but Clojure community also doesn't recommend writing macros unless you have to. There are also protocols and types, but they're not encouraged to poke around, it's still map-first recommended across the community.
There are very few language communities follow the "It's just data patterns". From Java to Python to ML, people mostly modeling things with fixed-property model, instead of just untyped maps, unless they have to.
For Lisps, I don't see map-centric approach is that stressed. Scheme might be still list-centric, Common Lisp is more diversed when it comes to approaches.
Why people hate JavaScript so much? And why people speak highly of Clojure? It's just about inconsistent details, or about the whole design itself? Is there any important point I missed?
Some people also find the lack of random syntactical noise in Lisp style syntax preferable.
It could be vastly improved with a proper REPL-based development (i.e. think developing within the browser dev tools, kinda, sorta), but instead we are now going the other way: Adding types and a compiler and making it harder to mis-type things, but also harder to work as productively.
I guess overall it has a few more ways on how you can shoot yourself in the foot compared to cljs ;)
I need to learn Web, as mucj as I don't want to, and love clojure and fp. Also figwheel looks awesome, but how hard is it to setup.
I also think maybe to really get the best out of clojure script you also need to learn react??
I'm not sure this is the case with elm.
Any other non js alternatives?
There’s almost no comparison.
Elm compiles much more quickly.
Elm produces much more performant code.
Elm enables you to write software that works reliably much more quickly because type errors happen at compile time, rather than at runtime like in ClojureScript.
ClojureScript’s error messages are at times incomprehensible. Elm’s error messages are best in class.
Elm is far easier to learn than ClojureScript, and you’re right — you do need to know React if you’re writing a non-trivial ClojureScript application.
Elm is criticised for not allowing a synchronous IO escape hatch. This is a pretty bad criticism I think. Being able to guarantee where effects won’t happen is a pretty great thing. Any effects work and/or JS interop can be done with ports. Ports work just fine.
Elm is also criticised for denying the user a lot of “power features” like typeclasses. I find it mildly annoying sometimes that I don’t have everything I want, but I understand the trade-off. This is to make the language more accessible to your average JavaScript developer who has never tried anything from this family of languages. And I sure as hell don’t find it annoying enough to want to drop the type-checking compiler.
You don't need to learn React to start using Re-agent or Re-frame.
IMO among all alt-js languages, Clojurescript is the most practical and production-ready.
I'm a desktop application programmer, not a mathematician. I don't need to print the first 25 squares of integers or Fibonacci whatever. In fact I think the hardest math I did at work was using modulo to get even and odd numbers ...
Show me how you parse a csv file, how do you connect to webservices or retrieve data from a database, show me how to do a simple GUI, show me something useful for the common dev.
I'll judge your ultimate programming language on these menial tasks that comprise 99.9% of my programming time.
The rub however is that GHC can check the consistency of my software faster than I or my colleagues will ever be able to, and programmer time is expensive.
So that's why not Clojure.
That's why you need to spend 3 years learning Haskell? I feel that's the right amount of years for an average Joe programmer to reach meaningful productivity with the language. Well, maybe I'm exaggerating, maybe it's just two years. Plus/minus another year to learn and understand compiler pragmas.
I’m not sure how that’s a rebuttal to my point though.
I run a startup. I’ve already put the time in to learn the technology. I’ve learned other technologies too. This one happens to be the one I find the cheapest (now) to write. Did it take my employees a long time to learn Haskell? Maybe! But I didn’t pay for that. They learned the technology for whatever reason, and I’m seeing far better returns on my investment than I think I would have if I hired a bunch of Clojure developers. My previous full-time gig was at a Clojure startup. Their pace is pretty atrocious compared to the pace of my current team.
At least for OCaml/ReasonML it was "less than a week" for me and Haskell is probably not that much harder than another ML.
For one thing, Clojure has type analysis: spec. Also, as pointed out in the article, you can also write tests.
+----------+-------+-------+
| Language | Tests | Types |
+----------+-------+-------+
| Haskell | Yes | Yes |
| Clojure | Yes | No |
+----------+-------+-------+What makes clojure exciting [for me] isn't the ability to build efficient transducers or create powerful custom syntax or build complicated value-level dispatch hierarchies (and in my experience, although they are cool and fun to use, all of these things certainly do make for code which is hard to understand and harder to debug), it's the comprehensive and easily-composable functional programming primitives, excellent built-in immutable data structures with intuitive shared interfaces, practical and (for the most part) straightforward approach to polymorphism, and sane approach to concurrency.
pure functions are completely separate to every other parts of your software, they are very readable and easy to debug
* edit: don't write functions longer than a few lines
I can't think of two more opposed camps in programming.
Edit: Joking aside, I do think you're right in terms of the ideological foundations and organizational structures of the two languages (clojure's emphasis on practicality, curation and evolution of the core library+language by a single BDFL, compared to haskell's origins in PL research and democratic design/maintainence), but in practice I think a lot of that comes out in the wash--dynamicism vs. static typing is definitely a huge difference, but there's a lot more to language design and the material practice of building software than whether your compiler reads type signatures, and I think that along those axes, especially when compared to other mainstream programing languages/ecosystems, clojure and haskell end up being pretty close.
One of my teammates also has production ClojureScript experience, and called it the worst of both worlds (Clojure and JavaScript). The primary problem is it doesn't try to abstract away the DOM, the most typically problematic part of working with JS.
Just my own encounter with it so far - I haven't had the experience of writing production Clojure myself, but from what I've seen with the syntax, I'm still happily primarily writing JS so far when I'm not writing Python/Scala/etc.
It does require a different way of thinking. I was lucky in that I had very little experience, academic or professional, when I picked it up. I had a lot less to unlearn.
But that doesn't mean it's limited to ultra-brainy lisp weenies. We have undergrads writing clojure, never having written clojure before, in their first programming job. They do just fine.
car and cdr are accessors for the fields of a basic data-structure; nothing academic about that. The very origin of the names is rooted in systems work, not academics.
(caddr x) provides a shorthand for (car (cdr (cdr x))), which is a pragmatic thing.
You know, like #(* % %) instead of (lambda (x) (* x x)).
The lines are more concise. That also means you will need more time per line to read the code.
Especially in Java, you can skip over so much boiler plate code, you kinda feel you are reading code fast.
Not so with Lisp. You need some familiarity with it and you will need to adapt to a line just having more stuff to understand in it.
Also, you can tune this using libraries as many things are a la carte.
So it's ok to say #(25 % %) because we both speak the same language and we know what it means, the intent is clear.
But it's not okay to call your variables a, b, c, because that's meaning, not syntax, and I can't understand your intent from those variables.
Massive one-liners probably fall into the latter, but I don't think he's advocating those.
(println (take 25 (map #(* % %) (range))))
⍪×⍨⍳25
Where is the evidence?
Back in the 1990's IBM taught many of their consultants Smalltalk, and researched what made learning Smalltalk difficult —
Smalltalk, although recognized as a good platform for rapid prototyping and software reuse, is widely regarded as difficult to learn. Unlike learning a procedural language like Pascal or C, learning Smalltalk is dominated by browsing and code comprehension. Learners of Smalltalk typically experience a long, slow start-up phase in which they become familiar with the class hierarchy and object-oriented computational model but do little meaningful work (our colleague Dave Smith calls this "climbing the Smalltalk mountain").
Programmers having to wrap their minds around the class hierarchy and OO was seen to be the problem.
public class SquaresOfIntegers {
public static void main(String[] args) {
for (int i=0; i<25; i++)
System.out.println(i*i);
}
}
First of all, Java is famously verbose. Nobody is impressed when a language is more concise than Java.
But even ignoring that, the comparison isn't representative. The enclosing class and 'main' function aren't needed for every additional fragment of code, so it probably shouldn't be counted here.
OCaml certainly has a lot of flaws, and is not nearly as "nice" a language as clojure, but the productivity of static typing (in the statically typed functional language sense, not the C++/Java sense) is huge. Knowing you can do a big refactor and the type system has your back is massive.
We sponsored core.typed to add types to Clojure, but there were flaws at the time (they have have been fixed since), and we didn't end up sticking with it.
So Clojure I wouldn't use again. OCaml I would, despite having significant flaws (every language has significant flaws).
Am I correct in summarizing your experience as:
The possible encumbrance caused by types at prototyping phase is paid back several times when refactoring a production codebase?
(>defn my-fun [a b] [int? -> int?])
It also intigrates Shaun's awesome Parinfer parentheses management system, which adjusts your parentheses according to your indentation. Think Python rules: Put a thing inside another thing by just tabbing it over.
So I think it does come down to if you like static types or not.
For scripting, there's a lot of ways to script efficiently with Clojure, it's actually great for it as well. So, for most my scripts, I don't really mind waiting 1 second for them to start. That said, if you wanted them to start faster, using ClojureScript takes you under 100ms. And using Joker (https://github.com/candid82/joker) takes you under 50ms.
When I need even faster startup for scripts and also need performant runtime, I've been using GraalVM to compile to native binaries, those start in under 10ms, and execute much faster than ClojureScript and Joker. That's a good strategy if you want to write things like a grep tool, or ls, or a linter, code formater, etc.
https://www.innoq.com/en/blog/native-clojure-and-graalvm/
Sometimes the trade-off of slow startup vs. having a JIT'ed VM is better, since you get a lot more performance, if you're working w/ long-lived processes.
Of all the things I’m unfortunate enough to have to manage in production, the JVM has never caused me any problems.
I mean, yeah, I’d like static binaries too (graal is promising but not quite there yet), but managing the JVM has nothing to do with it (footprint and startup time are my primary reasons)
This doesn't really exist in a lot of dynamic languages and especially in Common Lisp, Clojure, APL, Forth, Smalltalk...etc. You write a small piece of code and interactively run it with some data...test if for some scenarios or build actual tests if you want. When it is good, add it to your actual code and you're done. You interactively test everything. Once you get used to it, anything without a REPL is like wading through quicksand. The type error would probably be caught quickly when using the REPL. In Common Lisp, when your code errors out, it drops you into the REPL and you can patch your code while it is running. Smalltalk also has this. Again, when you've seen this, a Java app crashing seems hokey.
Otherwise the speed is almost identical to the speed of native Java and native Javascript. Except when you experience the power of lazy structures, then Clojure is faster than native code.
Yes, Lisp makes this a non issue for the most part, because the compile and test iteration loop is instantaneous and integrated fully within your IDE/Editor
Still, you get pretty decent auto-complete and static linting warns on quite a few things. For that though you'll want to use IntelliJ with Cursive or Emacs with Cider and clj-kondo + joker flywheel linters.
Edit: Let me address the performance part of your comment as well. I think using a game with 20fps as an example was to show that it could even achieve such performance. Languages like Java, C#, Clojure, Python, Ruby are normally bad choices for games as they are not performant enough. So most games are implemented in C++ with an embedded scripting language on top. So the fact a pure Clojure game can hit 20fps with lots of on screen object is actually pretty good in this case.
In general, for idiomatic Clojure code, you should expect to be within 10% the performance of pure Java. Non idiomatic Clojure can often match Java's performance, and when it can not, you can implement the hot paths in Java and use interop very easily.
For ClojureScript, performance is pretty on par with JavaScript.
Startup times are the biggest issue, if pure Java takes 80ms, Clojure will take more around 500ms to start. The issue is that each Clojure function is a Java class needing to be loaded at startup, and the JVM is very slow at loading classes. GraalVM can be used to make native images, and those will start in around 10ms, but your code might need to be adjusted a little as the native images don't yet support all runtime features.
ClojureScript startup times are pretty on par with pure JS running on Node.
What? A hundred on screen objects is not a lot, and 20fps is a slideshow. Pure Java games regularly do 10x that (Minecraft says hello as a simple example, and it's considered badly optimized)
A "proper" C/C++ engine you'd expect 100x (thousand objects at hundreds of FPS, or things like Factorio which are 10s of thousands of objects at 60 fps).
The anecdote performance is what I'd expect from a joke language like Rockstar. I'm sure the poor performance is not Closure's fault necessarily, and it probably can fly in the right circumstances (or maybe even in a lot of circumstances). Just the anecdote's description makes Closure sound horrifically slow.
It's not the job of the language, but rather that of the developer to write clear and readable code.
This problem is addressed by using good design practices, code reviews, testing, and documentation.
But at least the Java is overly verbose. It takes a while to read all that code. With the clojure stuff, I'm staring at the same line for just as long and getting no where.
I've inherited code bases in different languages in the past, and generally, it's the unfamiliarity to the language that makes it harder. Especially Clojure, being so different, the unfamiliarity is even stronger.
For example, if you read the Clojure source code (https://github.com/clojure/clojure) are you similarly lost? Or when reading any random Clojure github project?
Also, is this the first time you inherit software like that? Like have you ever inherited other similarly complex piece of code in your main language? Because I have as well, and it is never easy, no matter the language and familiarity.
Not trying to attack you by the way, it may be true, but I'm trying to isolate the variable of it being written in Clojure against the rest to really get a feel of the effect Clojure has over what you are experiencing.
All the rest, my team has managed to decode the intent and flow.
Do you believe the language is the issue, been written by someone not trained on software engineering practices?
Would you find untested (I'm betting this code you inherited lacks tests, from experience working w/ ML research code artifacts) verbose Algol-family code any easier to understand?
But we own lots of legacy stuff. It's only this one that gives me a headache.
Honestly, I wish it had bugs or failed occasionally so I could justify replacing it!
The names of complex functions, even with documentation, tends to be a bigger problem with scale than local variables and functions or utility functions.
Language design stuff like namespacing/modules helps provide a balance between short/descriptive and so does documentation of key/complex parts as well as how you architect the larger pieces and organize code in directories.
Additionally there's type systems, IDEs with docs/function + type signature integration, standards like JSDoc, etc that help. So it's a mix of a lot of things besides the human.
It is a very different approach and takes time to get used to.
I'm stuck on the other side: I have millions of lines of legacy Java code with high complexity. Even with all the types and very long names, understanding the code is not easy either and the 'step into/step over' buttons in the debugger are seared into my brain. Luckily IDEA can run arbitrary Java expressions during debugging, which gives you some interactivity back.
With Python, I find it's easy to understand what individual lines are doing, but really hard to understand what the codebase is attempting to do as a whole.
(->> (d/load-workbook file "Questionnaire.xlsx")
(d/select-sheet "Sheet1")
(d/select-columns {:A :item :B :price}))But here's CLojres read a file line by line and print it out:
C:\\foo.txt
"THis is line 1
This is line 2
This is line 3"
(with-open [rdr (io/reader "C:\\foo.txt")]
(doall (map println (line-seq rdr))))
=>THis is line 1
This is line 2
This is line 3
(defn read-file-line-by-line [file-path]
(with-open [rdr (io/reader file-path)]
(doall (line-seq rdr))))
(read-file-line-by-line "C:\\foo.txt")
=> ("THis is line 1" "This is line 2" "This is line 3")
(map clojure.string/upper-case (read-file-line-by-line "C:\\foo.txt"))
=> ("THIS IS LINE 1" "THIS IS LINE 2" "THIS IS LINE 3")Stolen from the README page for a clojure CSV parsing library (https://github.com/clojure/data.csv)
(defn read-column [reader column-index]
(let [data (read-csv reader)]
(map #(nth % column-index) data)))
(defn sum-second-column [filename]
(with-open [reader (io/reader filename)] ; Read in the CSV file (streaming / lazily)
(->> (read-column reader 1) ; Convert to just the first column of data.
(drop 1) ; Drop the first row (the CSV header)
(map #(Double/parseDouble %)) ; convert each string in this column into double
(reduce + 0)))) ; sum the result
(client/get "http://example.com/resources/3" {:accept :json :query-params {"q" "foo, bar"}})
It's a wiki, not a blog, but I've lost countless hours browsing through it and trying stuff out for myself.
Clojurescript has heavily bought into React. No cljs programmer I know deals with the DOM, but rather uses a React-based library to do it. Personally, I use re-frame and reagent and haven’t needed to interact with the DOM myself in years. I find working in cljs a LOT more productive and pleasant than working in Javascript (which I also use a lot, although if given the chouce I’d use cljs instead).
OP said no wrong, but think about it: Bitcoin also just stole the good parts from other software projects.
Really practical inventions usually gather the best ideas from a field.
Uncle Bob didn't mention any spec about his computer, he didn't say if it's 2D rendered or 3D rendered, if he's using GPU acceleration or not. So I had a look at the code for his game, and it is using a software 2D renderer. That means no GPU acceleration, it is fully rendered on the CPU. It basically uses the default processing.org 2D renderer.
Now, I don't know what spec his computer has, but software only 2D renderers are generally pretty slow. So 20fps isn't bad.
Anyways, I agree it's a weird way to show a language's performance. Like I said, Clojure will run around 10% slower than pure Java in most cases. Which is pretty fast, since Java is arguably the most performant GC language around.
For example if I see something like (filter even? numbers) I immediately know what is being done and why. There aren't any surprises there.
Meanwhile, when I see something like:
List<Integer> acc = new ArrayList<Integer>();
for (int number : numbers) {
if (number % 2 == 0) acc.add(number);
}
So, there's a lot more to think about even in a completely trivial case such as the example here. In real world code that complexity quickly becomes overwhelming in my experience.
Doing it in Python is easy...it is a bigger pain in several of the other ones. Even your Clojure examples are definitely more work than in Python. You even defined a helper function (something I've never needed to do in Python) as the default is so easy. Mind you, the Clojure isn't too bad and definitely better than some of the others I've mentioned.
To your average developer...this is indeed probably irrelevant as IO might be a small part of your codebase. I do a lot of process automation though and do a lot of IO.
Here's a few examples stolen from the doc page for the next.jdbc library (https://github.com/seancorfield/next-jdbc):
> clj
Clojure 1.10.1
user=> (require '[next.jdbc :as jdbc])
nil
user=> (def db {:dbtype "h2" :dbname "example"})
#'user/db
user=> (def ds (jdbc/get-datasource db))
#'user/ds
user=> (jdbc/execute! ds ["
create table address (
id int auto_increment primary key,
name varchar(32),
email varchar(255)
)"])
[#:next.jdbc{:update-count 0}]
user=> (jdbc/execute! ds ["
insert into address(name,email)
values('John Smith','john.smith@email.org')"])
[#:next.jdbc{:update-count 1}]
user=> (jdbc/execute! ds ["select * from address"])
[#:ADDRESS{:ID 1, :NAME "John Smith", :EMAIL "john.smith@email.org"}]
You can see the return from the select contains clojure native data-structures. It returns a list of rows, where each row is a map of fieldName to fieldValue. You can now go to town and manipulate lists and maps to your hearts content, convert the payload to HTML or JSON or whatever you like.
But your simple dismissal of Clojure (even though seem to be an educated one) has no merits. Don't feel bitter because Clojure is growing in popularity (in relative terms) and Haskell seems stagnated. Haskell is an older language and has had its ups and downs, and I think it will be fine. Clojure as well - no matter how aggressively haskellers would evangelize against it, it will continue to grow steadily. Think of it as a gateway drug into Haskell. I know people who jumped that way, and I know people who went the other way - from Haskell to Clojure.
To me, it is like using headphones - some prefer wired ones, some people enjoy wireless, even though audio quality suffers, wireless gives them a lot of freedom and might be viewed as a "more practical choice." To that though, proponents of wired might say: "what if the battery dies?". It is an endless and quite pointless debate. So let's concur: Haskell is great, and Clojure is fantastic too.
I believe it has an impact on my business' bottom-line. Saving money in business is a merit.
> Don't feel bitter because Clojure is growing in popularity and Haskell seems stagnated.
That… was an odd take. I'm not sure what gave you the impression that I would be bitter about Clojure growing in popularity (not that I think it is, anyway). I have nothing invested in Clojure not being a popular language, and the idea that Haskell has stagnated is a little hilarious.
> Haskell is great, and Clojure is fantastic too.
Again, I have never denied Clojure's merits. My position is pretty simple — I have found it cheaper to build Internet software businesses with Haskell than with Clojure.
OCaml was because I had experience with Haskell and didn't like it, and with Elm, which I did like. We tried it for a while as an experiment and it worked really well.
But for backend, the benefits of static sound attractive to me only on a program/product that is dominated by its internal business logic and has less interfaces to the unpredictable and constantly changing outside world. I think Darklang may be like this, from looking at the medium.com link?
It becomes unwieldy like every other language.
> What happens if Questionnaire.xlsx or Sheet1 doesn't exist ?
Runtime exception.
I've seen too much Clojure code (and this goes for every language where error handling is optional - Hi Javascript!) just skip error handling and focus on "happy path".
And as demonised as Java's checked exceptions are, they at least forced error handling to be thought of.
Ya'll underestimate the power of a programming language to create a programming language. Just replace the arrow with `ok->>`
Or is there many things to consider at once in the system instead of many individual things?
I do think that a static analysis tool would be of some help. Some sort of tooling to better handle tree structures would be very handy. I often find myself being off-by-one level with get-in calls on tree data (E.g. (get-in m [:a :b :d]) where m is {:a {:b {:c {:d 1}}}}), which is annoying because the NPE gets thrown 3 function calls up the stack.
The last Clojure team I worked with decided to add “rop” to our project. The documentation in this library seems to go to great lengths to avoid using the word “monad”, despite the fact that the entire idea being encouraged by this library is monadic composition, but specialised to the Either monad. I mean come on, this library even takes its bind operator from Haskell.
These Clojure programmers were not Haskellers. They were just as indoctrinated as any other Clojurists and probably would have made the same flippant comment that you did.
And yet, here we are.
> clojure and haskell end up being pretty close.
In terms of the cost of building a software business, this has not been my experience. In my experience, the cost of building software at a scale beyond trivial is higher with Clojure than it is with Haskell.
First, to be clear, haskell's type system is powerful and interesting and useful, and clojure's lack of good static analysis tools IMO represents a serious [comparative] deficiency in the ecosystem (and spec is obviously no substitute). I don't think that it will ever be possible for clojure to get even close to the sort of compile-time checking/guarantees that haskell offers, nor do I think clojure is particularly well-suited for the powerful higher-level abstractions that you see in e.g. lens or recursion-schemes (or, honestly, even Control.Traversable). My personal experience (both as a hobbyist and professionally) has been that there are some practical advantages from clojure's dynamism such that it probably makes sense (in a cost/benefit sense) to eschew static typing in some/many circumstances, but I really only have anecdotal evidence plus some intuition backing that up.
That aside, the point I was trying to get at was that both haskell and clojure encourage you to build complex data representations out of simple, easy to understand, composable, persistent structures, and then build up computations by composing pure transformations on them. Additionally, both languages have reasonable concurrency models/primitives, good tools for creating expressive DSLs (if that's your thing), and powerful interactive development environments. I, personally, find all of these features to be incredibly important for effectively developing software--they shape in a very real way the processes I use to plan and write and debug code--and most languages in common use today lack at least a couple of them (some languages have none of them, in fact).
Frankly, it’s not often I hear a rational argument like this in a typical Haskell vs Clojure debate. Usually I just hear zealotry and a refusal to address Clojure’s shortcomings, just like in the Robert Martin blog post. In fact come to think of it, many of the other comments on this thread which are replies to criticisms of Clojure are along the lines of “well, maybe you just didn’t try hard enough. Maybe you failed. It’s not the system that failed. Clojure is too perfect to fail.”
This kind of rhetoric is frighteningly similar to that used in multi-level marketing.
I won’t deny Clojure’s benefits. They are there, and I did say in another comment that Clojure is one of the most sound dynamic languages. But I don’t think everyone is honest like you were just now about its drawbacks. Robert Martin certainly isn’t honest about it.
Check out the success stories on the Clojure site. One of the biggest ones to me is Walmart using it for Black Friday e-commerce. That is a good indicator to me. They've had good success stories with F# as well, but C# uses F# as a research language and the best bits are kind of bolted back onto C#. It keeps C# as less painful than Java, so less devs make the jump. Honestly, the future for both languages isn't 100% secure. I just wish both languages got more love.
I'm in the same boat as you. Clojure seems to really be meant for someone who already has a few years of Java experience. There are plenty of examples where someone dips into Java and I'm totally lost.
Let's talk about writing code in Python. You write a script and either run it through your IDE or via the command line via Python pythonscript.py. With D, I write some code and feed it through the compiler like dmd scriptname.d.
With Clojure in addition to all the cruft I have to learn (Emacs, leinegen...etc), I still have to know something of Java. One of the big selling points is integration. If I know zero Java, how am I supposed to make use of the various classes? Even if I largely stick to just Clojure, how does one interpret the Java stack traces? What is the best method for running via the JVM? How do I performance tune the JVM? How do I use GUI frameworks that are largely all OO Java?
And yet... I agree with GP. I love clojure, but over time I’m becoming less and less sold as dynamic typing. Spec helps, a little. Property-based generative tests (especially when used with spec) helps a little too. Neither are a replacement for proper static types, though, especially an ML-esque type system with type inference. Bonus points if spec validation could pass type data to the compiler/type inference (eg the code path after validation can assume that the data is of the the types described in the spec).
I dream of a statically typed Clojure with type inference, spec-inferred types and optional dynamic typing for REPL experimentation and glue code.
Could I ask what kind of project it is? I wonder if I'm just lucky that Clojure fits perfectly my use case, which is mostly a set of distributed systems of all kinds. So while as a whole there's tens of thousands of LOC. The components have very strong boundaries being as it's a set of services assembled together through RPC, PubSubs and DBs. Maybe that alleviate the lack of a static type checker.
We've adopted Clojure about 3 years ago, team of 10. We've had a few people leave and join throughout. Only one person knew Clojure beforehand. Our stack is about 50% Clojure, 40% Java and 10% Scala. Of all three, Clojure has given us the least issues, has been pretty easy to maintain and generally has fewer defects. Java tend to have the most bugs, almost always related to some shared state. Scala I find the hardest to extend and maintain, but the code base we have for it I think does Scala wrong, it's like the worst mix of OOP and FP.
> I dream of a statically typed Clojure with type inference, spec-inferred types and optional dynamic typing for REPL experimentation and glue code
That's pretty much exactly core.typed: https://blog.ambrosebs.com/2018/09/20/towards-typed-clj.html
That said, the project never managed to get more contributors.
If you're looking for a typed Lisp, I've been keeping my eyes out on Carp: https://github.com/carp-lang/Carp
Don’t get me wrong, if I restarted the project again from scratch, I’d choose the same setup (or a very similar one). I love Clojure and am very productive in it, but that doesn’t mean I don’t think it could be better still.
> That's pretty much exactly core.typed
I haven’t looked at it in a couple of years, maybe its changed, but when I did, it didn’t really do it for me. It’s still a separate tool that lives separate from Clojure itself and it felt very “heavy”. In my personal opinion and experience, having certain things as separate entities (decomplected as Rich would say) isn’t always a good thing and leads to an inferior thing. I’ve played around with many programming languages in my time (I’m a bit of an enthusiast, I guess. I like trying out languages that are very different from what I already know – that’s how I originally got into Clojure) and it seems like a common theme. The closer a feature is to the compiler/runtime, the better it works and seamless it is over all. Another Clojure example is the limitations core.async has, because its an external library: things like <! cannot be placed inside functions or the go macro can’t see it to transform it as macros cannot look inside function calls. I’ve also encountered an exception recently where the stacktrace only showed core.async and clojure.core code, not a single stack frame referenced MY source files. These problems are hard to solve as an external library.
> If you’re looking for a typed Lisp
I’m not. I like Clojure’s particular mix of sensible syntax, immutability, sequence abstraction and general way of doing things. Other Lisps I’ve looked at don’t have the same emphasis on these things as Clojure does, so I don’t want another Lisp, I want a language that makes the exact same decisions and tradeoffs as Clojure, except on dynamic vs static types (and actually useful error messages). Maybe one day I’ll give it a try, I certainly don’t expect Cognitect to change their language because of what my preferences are.
import really_long_name as rln
rln.foo()I understand the idealist world view is seductive — we'll be discipined, we'll write the tests, we'll make good use of Spec, etc etc.
In reality, I have never seen this happen. Humans are lazy and undisciplined, so that's what I optimise for.
It’s not a type system, but it’s very concise and easy to add validations to important functions. This removes the “lazy and undisciplined” part of the equation.
I agree.
But we do use spec very successfully for complex data structure validations and example data generation.
This is also made obvious by the idiomatic use of each technology.
So, no. My little chart is correct.
There are downsides, however. The UIs that it creates definitely look like Swing apps. I.e. ugly at least on Linux. Also, the up to date documentation is difficult to find, I'll like to it here when I get home. The GitHub-linked docs work fine, but they do miss a few features.
Check out the examples/mini-tutorial: https://github.com/cljfx/cljfx
It's all React-like. You have a state atom and a GUI map data structure (which I think is equivalent to your DOM in React..) and then you hook up events that update the state and blah blah. It's all very clean and easy to read/use. The underlying JavaFX is also great for an OO GUI library so it's not gross to dive into if you need it
As a general first rule, make the alias the same as the namespace name with the leading parts removed.
(ns com.example.application
(:require
[clojure.java.io :as io]
[clojure.string :as string]))
[clojure.data.xml :as data.xml]
[clojure.xml :as xml]
[clj-http :as http]
[clj-time.core :as time]
[clj-time.format :as time.format]
Namely:
(require '[dk.ative.docjure.spreadsheet :as spreadsheet])
About `d`:
There are always exceptions. For example, some namespaces have established conventions for aliases:
[datomic.api :as d]The issue is how would type definitions be introduced, and what kind of types would be most appropriate? As you said yourself, the way core.typed did it felt too "heavy". Yet it isn't clear how to make a more lightweight variant for a Lisp language such as Clojure without rendering the types worthless.
The first, and one of the biggest issue in my mind, is that what everyone loves about Clojure is the data-oriented style. In that style, you represents entities and their relationships using heterogeneous collections. That's where in Clojure you model your domain with Maps, Lists, Vectors, Sets, etc. It is awesome, but no one has figured out a non "heavy" way to statically type it. All methods I know of have bad programmer ergonomics. In effect, adding types back to it almost kills the data-oriented style, and it ends up feeling a lot like modeling with Classes instead. See Haskell's wiki section on this problem: https://wiki.haskell.org/Heterogenous_collections they haven't solved it, and have multiple ways to possibly handle the scenario, and non are ideal.
> I want a language that makes the exact same decisions and tradeoffs as Clojure, except on dynamic vs static types
I would too, but with the caveat that the development experience would be the same, and the programming ergonomics and styles would be retained. And this, I'm afraid, is an open problem that no one has solved yet. It's not just a case of personal preference. Having a language which has the pros of Clojure and the pros of static types, without the cons of static types is hard. That's why for now, you need to choose one or the other.
Take the following clojure:
(+ 1 2)
=> 3
It means if we do this
(1 + 2)
However, we can tell Clojure to treat this code as data by quoting the list using the '
'(1 + 2)
=> (1 + 2)
WHat if we write a function to swap the first two items in a list?
(defn swap [x]
(list (second x) (first x) (last x)))
(swap '(1 + 2))
=> (+ 1 2)
So you can see
(+ 1 2)
We can eval it:
(eval (swap '(1 + 2)))
=> 3
Up steps macros, macros are evaluated before compile time and don't evaluate their arguments. So we can rewrite swap as a macro
(defmacro swap [x]
(list (second x) (first x) (last x)))
(swap (1 + 2))
=> 3
swap = (x,y,...rest) => [y,x,...rest]At that point you're emulating lisp without the elegance, and the first approach is only possible because functions are first class objects. If you were trying to rearrange an expression that had control flow or keywords in it (eg. Modify the behaviour of an if) then you would have to reify the "if" (change the original code so it had a class to represent the if not use the keyword). So you're kind of reinventing lisp by wrapping every part of your language as an object
The main point of macros is when you use a regular function in Clojure you have applicative order evaluation. Macros do not, as macros are designed to transform and generate code.
A better example would have been the (when) macro.
In Clojure you have (cond) and (if) for conditional evaluations. If has the form
(if (cond)
(when-true)
(when-false))
(if (= 1 1)
(println "1 = 1")
(println "uh-oh the unviverse is broken!"))
=> "1 = 1"
(if (= 1 1)
(println "1 = 1")
(println "also hello"))
(if (= 1 1)
(do
(println "1=1")
(println "Also hello")))
1=1
Also hello
(when (= 1 1)
(println "1=1")
(println "Also hello"))
"1=1"
"Also hello"
Essentially, we want to write
(when (= 1 1)
(println "1=1")
(println "Also hello"))
(if (= 1 1)
(do
(println "1=1")
(println "Also hello")))
This is incredibly easy in Clojure. We just need to create a list where the first item in the list is `if`, the second item is the conditional we pass to the macro, the 3rd item is a sublist, of which its first item is `do` followed by the list / functions to execute when our test evaluates to true!
(defmacro when [test & body]
(list 'if test (cons 'do body)))
If we run the macro expansion on when we can see the code that it generates:
(macroexpand '(when (= 1 1)
(println "1=1")
(println "Also hello")))
=> (if (= 1 1) (do (println "1=1") (println "Also hello")))
Say we have the nested function calls:
(reduce + (filter even? (range 1 11)))
=> 30
(->> (range 1 11)
(filter even?)
(reduce +))
(defmacro ->>
[x & forms]
(loop [x x, forms forms]
(if forms
(let [form (first forms)
threaded (if (seq? form)
(with-meta `(~(first form) ~@(next form) ~x) (meta form))
(list form x))]
(recur threaded (next forms)))
x)))
Basically, using macros you do things like control symbolic resolution time, extend the compiler to create a DSL spcific to your domain and reduce boiler plate code.
That's before you start getting into properly weird stuff like anaphoric macros.
Macro arguments are not evaluated, and so this works.
You could make it a function, and pass the arguments quoted, but it'd be more cumbersome.
Macros are actually my least favourite part of Clojure. Sure, its great to have them and there are some libraries that use them to excellent effect (instaparse, Hugsql, etc), but most of the time, I prefer tools that don't use macros. You can see by comparing libraries that were made in the earlier days of Clojure versus more recent ones: the earlier ones love to use macros while the newer ones prefer functions and data. The data-first libraries are, in my opinion, easier to test and easier to build complex things on top of. If I look at a readme and see that the expected way to interact with a library is a macro, I usually look for a data-first library instead and only use the macro one if I can't find an alternative.
Macros are technically cool, you can do some interesting things with them, but I find that in most cases where they're used, they are inferior to alternatives. Its technically very cool that core.async could be implemented as macros, but I feel that core.async greatly suffers from it versus being built into the runtime: you cannot use many operations in functions called by core.async/go because the macro can't see inside function calls. Also, I've had exceptions thrown by core.async where the stack trace did not mention any of my code. That was not fun to debug.
Don't get me wrong, macros do have their place: hugsql uses macros to parse the SQL file and generate functions for you, which is awesome! But for every great macro-based library, there are many more that I wish didn't use macros.
(This is user-facing macros in a libraries' API. I have no problem with using macros internally)
I do love Clojure overall, though.
But having that option is better than not having the option at all IMHO.
When I meet a girl, I ask her about macros and functions first.
I except to hear something like: macro < function < data.
I've read a few Lisp books and dozens of internet blog posts, so I know about macros and why people use them without getting the full understanding which comes with actually writing code.
cljs.user=> (defmacro swap [x] (list (second x) (first x) (last x)))
#'cljs.user/swap
cljs.user=> (swap (1 + 2))
Execution error (Error) at (<cljs repl>:1).
1.call is not a functionThanks for helping point this out. Cursive has a trial license and is pretty cheap, but it is still annoying to get a license to something you might not need in my organization if others aren't already using it (probably easier to get visual studio for over $1k than $100 cursive license). VS Code asked me for the leiningen REPL path and I just gave up at that point for the night. I wish the Clojure download zip came with a very simple IDE like Dr. Racket or the GroovyConsole with line wrap, parenthesis matching/highlighting and a pane for the file and REPL and some pre- loaded tutorials. I would have no excuses.
https://clojure.org/guides/getting_started#_other_ways_to_ru...
You can also use boot, or if you’re using clojurescript, there are other non-leiningen options too. But really, leiningen is easy — you only need to know where in the project file to put dependencies (if you’ve ever used any other build tool, you can do this) and then run `lein repl` or `lein run` or `lein test` and to package a jar: `lein uberjar`. You don’t need to know anything else to get started.
Cursive is free for non-commercial use. If you just want a simple editor to get started, check out Nightcode or Nightlight.
(defun swap (x y) (y x))
Then you can call it like:
(swap (2 3)) => (3 2)
(defn swap [x]
(list (second x) (first x) (last x)))
(swap (1 + 2))
=>Exception! 1 isn't a function.
So we quoted it in the function call by putting ' in front of the list '(1 + 2):
(swap '(1 + 2))
=> (+ 1 2)
So if I was to
(+ (swap '(1 + 2)) (swap '(3 + 4)))
=> Crashes! Can't convert alist to a number.
(+ '(+ 1 2) '(+ 3 4))
(+ (swap (1 + 2)) (swap (3 + 4)))
=> 10
(swap (1 + 2))
(+ 1 2)
You could do `(defn swap [the-list] ((second the-list) (first the-list)))` (which would invoke the second item as a function on the first item). It comes down to is the list a literal list as a parameter to something or is the first item a function to be invoked.