Clojurians Log v2

Cool, I've been using in #eca and clojure-lsp the java https://github.com/java-diff-utils/java-diff-utils, maybe that can replace it

@ericdallo sure, let me know if you need anything. I just implemented what I needed to get cljfmt's diff working in bb

@borkdude so this means we will see cljfmt running in bb natively soon? 👀

@ovidiu.stoica1094 Yes, already possible, but just the "check" functionality didn't work due to the diff lib. https://github.com/weavejester/cljfmt/pull/404

Other improvements • https://github.com/babashka/babashka/issues/1954: add clojure.lang.EdnReader$ReaderException • https://github.com/babashka/babashka/issues/1951: fix --prepare flag skipping next token • Document -D in bb help https://github.com/babashka/babashka/issues/1903 • SCI: fix .sym on user-defined vars returning qualified symbol instead of unqualified • SCI: add .ns and .sym field access on vars for compliment compatibility • SCI: ns-aliases now returns identical namespace objects as find-ns • SCI: alias now accepts namespace objects in addition to symbols • SCI: copy-var and copy-var* now preserve :private metadata • SCI: private vars are no longer included in ns-refers • SCI: deftype and defrecord now macroexpand to deftype*, matching JVM Clojure, enabling code walkers like riddley • SCI: case now macroexpands to JVM-compatible case* format, enabling tools like riddley and cloverage • SCI: fix .getName on custom types to return fully qualified name, matching Class.getName() • SCI: fix incorrect type hint causing method resolution failure when hint doesn't match runtime class • SCI: fix NPE in resolve when called at runtime (e.g. inside macro body during code walking) • SCI: fix .method on class objects routing to static instead of instance method path • SCI: macroexpand-1 now expands (ClassName. args) to (new ClassName args) and (.method ClassName) wraps class targets in identity, matching JVM Clojure • SCI: store current analysis context during macro invocation, enabling tools like riddley • Add clerk and cloverage to lib tests • Add ref-max-history, ref-min-history, ref-history-count • Hide redundant (keyword) description in REPL completions https://github.com/babashka/babashka/issues/1939: Disable JLine backslash escaping/shell history commands • Add various Java classes: java.lang.reflect.Constructor, java.lang.reflect.Executable, java.util.stream.Collectors, java.lang.StackOverflowError, clojure.lang.IType, java.util.Comparator (reify) • Bump rewrite-clj to 1.2.54, tools.cli to 1.4.256, transit-clj to 1.1.357, fs to 0.5.32

Thank you for supporting org.clojure/tools.deps.edn in 1.12.217! gratitude

I thought we stick to unix cli convention? This "-version" could cause confusion that may interpreted as seven single-character flags v e r s i o n

I know, it's super ugly but compatibility with clojure is one thing bb aims for

A nod to Java. 😃

And a nod to @seancorfield who asked for it ;)

I don't think we need to bend the rules just because "clj" or "java" has it. I'm sure (java/clj)'s cli design has its own legacy and history.

https://github.com/babashka/babashka/issues/1955#issuecomment-4049365324

@snow40479 clojure -version java -version cljr -version -- it's a pragmatic choice to be compatible with other tools in our (Clojure) ecosystem.

Well, given that we have things like '-Sforce' '-Sdeps', I guess the rule is already bent in bb, so never mind.

Java has supported --version for jonk's tho.

> java --version
openjdk 25.0.2 2026-01-20 LTS
OpenJDK Runtime Environment Temurin-25.0.2+10 (build 25.0.2+10-LTS)
OpenJDK 64-Bit Server VM Temurin-25.0.2+10 (build 25.0.2+10-LTS, mixed mode, sharing)

Yes, java -version and java --version both work. clojure -version and clojure --version both work too. Same with cljr.

The clj style of "-Ttools -M:xx -T:tool" is absolutely horrible, but I guess the virus has already affected bb, it's too late now.🥹

Try bb -XX:PrintFlags= for more exceptions to the beautiful unix cmd line rules ;)

(this comes from graalvm, not bb itself)

In a perfect land, I suppose we shall have some like --substrate-vm-opts="-Xmx256M" instead 🙂

graalvm doesn't let you handle these, it's too late by the time you get to the cmd line args

I see..

@borkdude it just occurs to me maybe there IS a way to enforce beautiful unix cmd line rules, but totally not worth the effort I guess. For example, if user needs graalvm flags and input bb --substrate-vm-opts="-Xmx256M"`` , we parse it and spin up a subprocess bb with the origianl command like bb -Xmx256M xxx that user won't notice. 🤔

@snow40479 And for bb -version it would spin up a new bb --version process? 😄

no, i'm talking about 'graalvm doesn't let you handle these, it's too late by the time you get to the cmd line args'. bb -version shall just get deprecated! haha

I don't try to hide or adapt graalvm or clojure CLI stuff, I embrace it. hiding stuff isn't the clojure way

yeah, it's just my random thought anyway.

Fun trivia: java -version writes to stderr, java --version writes to stdout.

if f throws fut never gets completed, unsure what the contract around exceptions futurize is meant to have, but that is still a change in behavior vs. a future from an executor service. another change is around future-cancel

can you put that on http://ask.clojure.org ?

@hiredman Regarding the contract around exceptions, we could define that to see if we could derive one. At the moment we only promise a Future but could say more if needed. Regarding, future-cancel can you elaborate?

I was thinking that completablefuture didn't have a .cancel but it does

I didn't look too much into yesterday, but that fact that futurize is exposed here https://github.com/clojure/core.async/blob/master/src/main/clojure/clojure/core/async/flow.clj#L329-L341 seems like it has preserve the exception throwing behavior on deref? otherwise any users would get a nasty surprise if their function throws and then the future never completes

and for example https://github.com/clojure/core.async/blob/master/src/main/clojure/clojure/core/async/flow/impl.clj#L259 could just hang forever if the step function threw for some reason

https://ask.clojure.org/index.php/14994/is-the-recent-change-to-flows-futurize-safe