This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
- # aleph (14)
- # announcements (2)
- # babashka (8)
- # bangalore-clj (2)
- # beginners (66)
- # calva (8)
- # cider (1)
- # clj-kondo (24)
- # cljdoc (3)
- # cljs-dev (3)
- # cljsrn (2)
- # clojure (197)
- # clojure-europe (1)
- # clojure-india (5)
- # clojure-italy (4)
- # clojure-nl (27)
- # clojure-uk (18)
- # clojurescript (56)
- # code-reviews (19)
- # core-async (86)
- # cursive (16)
- # data-science (1)
- # datomic (16)
- # docker (3)
- # events (1)
- # fulcro (101)
- # graalvm (10)
- # graphql (16)
- # jobs (1)
- # jobs-discuss (6)
- # kaocha (4)
- # luminus (1)
- # off-topic (99)
- # onyx (3)
- # pathom (9)
- # planck (2)
- # re-frame (8)
- # reagent (3)
- # reitit (3)
- # remote-jobs (3)
- # shadow-cljs (21)
- # test-check (3)
- # tools-deps (21)
- # vim (16)
just curious - what is the reasoning to have both Zulip and Slack channels? why not merge to one?
One reason ZulipChat Clojurians was created was because it has unlimited searchable history, for no $ cost. Slack's history for $0 is limited to the last 10K messages or so.
Which of these, or other, on-line forums people use is up to their own individual decisions. There is no entity commanding everyone to use one or the other.
I find Zulip significantly nicer, with a way better story for managing (navigating, reading) multiple channels and threads. I actually consume clojurians' slack from Zulip (and come here mostly to write)
Some context/rationale was can be found on Clojureverse: https://clojureverse.org/search?q=zulip TL;dr because of the limited retention, Slack is less suitable to host one's OSS-project or local meetup group, as channels with less traffic will look abandoned quickly.
awesome. so where should I ask my clojure-ish questions? 🙂 one-life-time with so many options...
on Slack there's way more folks, but that has the side-effect that your question may 'disappear' quickly.... The topics on Zulip make a question stand out more, but alas, getting an answer might take some time due to less visitors.
Right now, there are more people participating in discussion on Slack than on ZulipChat for Clojure, so you are probably more likely to get answers sooner on Slack. The discussions of most Slack channels is recorded on ZulipChat, where they can be searched over a longer history
it's also why we have multiple linux distributions or cars of different colors :man-shrugging:
Hi, as anyone used the clj-http client with an API that requires a client SSL cert? The term doesn't seem to be mentioned in the documentation. It would be the equivalent of the --cert option in curl.
Why is a function definition with
fn referred to as an anonymous function, when the
fn signature does include a name. Is it referred to as anonymous as it can be used without a name or because it is often used without a name? Or is it referred to as anonymous because it is not bound to a namespace scoped var.
Is it correct to refer to an
fn definition as a lambda function or an inline function (as its called inline as part of an expression/form).
Many people use
fn without a name. It is nice for easier understanding of stack traces if you do use a name for them, although as you say, those names are not callable from a 'global' context, i.e. at the top level of your code.
Historically, I believe the name anonymous function for the similar Common Lisp and Emacs Lisp lambda expressions (which used the symbol
lambda instead of Clojure's
fn ) was in common practice before Clojure was created. AFAIK, they did not have an option to give a name, the way Clojure
Emacs Lisp manual reference, where they are called anonymous functions: https://www.gnu.org/software/emacs/manual/html_node/elisp/Anonymous-Functions.html
Also Racket manual (derived from Scheme): https://docs.racket-lang.org/plait/lambda-tutorial.html
Thank you, this is very useful. I will be learning Emacs Lisp properly this year, so useful to contrast with Clojure.
I thought it was a general term used across P-langs as a synonym of
lambda expression. The definition being that it's not bound to anything.
Often you can take the return value of such a lambda expression and store it as part of some other data structure, e.g. an element of a list, or a value of a dictionary/map/etc. Not sure if that is what you meant by not bound to anything.
And yes, there is a Wikipedia page for "anonymous function" that lists a bunch of programming languages that have the idea: https://en.wikipedia.org/wiki/Anonymous_function
It mentions Lisp having the idea in 1958, but I am not certain whether the name "anonymous function" was used at that time yet.
right. I think "anonymous function" has that adjective because there are programming languages where you cannot make a function that is nameless.
C has function pointers, but I think a difference with anonymous functions (besides that they cannot be closures) is that they must all have names.
Very useful discussion, thank you. This gives me a lot more context. So my understanding is that although you can use a name with fn, that name is not available outside of the scope of the function itself. Therefore the expression is anonymous as there is no name for code outside of the function scope to refer to it. Including a name in an fun definition is useful for clarifying stack traces and making recursive functions (functions that call themselves). You can bind the function to a global name using def (or the defn macro) or bind the result to a local name using let. Thanks again, very useful discussion.
You can definitely bind the function to a global name using def / def, or to a local name using let, but in general, as for most languages that have functions as values, you can store functions as values anywhere you can store other values (e.g. in a list, in a vector, in a map, set, etc.)
Another definition of expression is that it's the return value you care about.
(fn ,,,) returns a function, without side effects.
fn still returns something -- is an expression -- when the function is named.
Hi, another beginner question: I'm having trouble figuring out some java interop. I have an ssl certificate and I'm trying to read the subject (so that I can parse out the email), but I've only gotten this far and can't figure out what to do with an instance of "java.util.Collections". This assumes (import 'java.security.KeyStore):
(with-open [cert (
(let [ks (. KeyStore getInstance "PKCS12")
c (. ks load cert (char-array "client"))]
(. ks aliases)))
How do I see that list of aliases?
Or perhaps someone knows of a clojure library that does this, so I don't have to work this low-level.
in clojurescript, i'd like to reload an audio element after changing its source...
is this a question about literally translating that to interop syntax, or about how to do that operation in some cljs framework like reagent?
or a translation that looks more direct but does the same thing:
(-> document (.getElementById "myAudio") (.load))
I've started using
-> a lot, because of its consistency between interop and Clojure usage:
(-> obj .val .val .val) (-> map :val :val :val) (-> function fn2 fn3 fn4)
So I actually use it for getting keys in map as well, even if only one, and I never need to use get-in in that case, since -> naturally extends as well
(-> map key) ;; instead of (:key map) (-> map :key1 :key2 :key3) ;; instead of (get-in map [:key1 :key2 :key3])
And it can combine between all three:
(-> (obj-returning-fn) .value (map-returning-fn) :user :name)
I think the reason you (and many others) are tripped up on translating that sort of thing is because it is chained method calls you parse it as a single thing, but if you break it up in to step by step method calls then the answer is pretty clear
If I am getting a GUID byte-array, is there something in Clojure/Java that will turn that into a dashed string? If I call
str on the byte array itself I get weird characters.
(fixed - leaving out encoding is deprecated)
user=> (String. (byte-array [104 101 108 108 111]) "UTF-8") "hello"
lazy first cut at what I think you want
user=> (map #(format "%2x" %) (byte-array [104 101 108 108 111])) ("68" "65" "6c" "6c" "6f")
@bfabry oh right, but make that "%02x" because leading zeroes are idiomatic I think
I feel like there's probably a cleverer/neater way to do it than converting it into a seq and joining it all up that way, maybe in a java guid class? but it works
here's the neat way if you're "embracing the host"
user=> (import 'java.util.UUID) java.util.UUID user=> (.toString (UUID/nameUUIDFromBytes (byte-array [1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16]))) "190c4c10-5786-3212-9d85-018939108a6c"
(cmd)user=> (java.util.UUID/randomUUID) #uuid "146dd4f2-35f0-4f2c-b134-7043d012ed27" (ins)user=> [(.getMostSignificantBits *1) (.getLeastSignificantBits *1)] [1472066789972135724 -5677789793651331801] (ins)user=> (java.util.UUID. (first *1) (last *1)) #uuid "146dd4f2-35f0-4f2c-b134-7043d012ed27"
I've never used name based uuids, but I guess that's what's going on if you get a byte array