Fork me on GitHub
#cljdoc
<
2021-11-19
>
martinklepsch10:11:41

That’s definitely a bug, but interesting

Jakub Holý (HolyJak)11:11:43

Good! I hate boring bugs 🙂

Jakub Holý (HolyJak)10:11:49

Q2: Does cljdoc support libs not publish to a jar repo, such as https://github.com/borkdude/grasp ? If so - how do I get it to display its docs? 🙏

martinklepsch12:11:19

Unfortunately that’s not supported. There’s an issue with more context here https://github.com/cljdoc/cljdoc/issues/459

👍 1
lread12:11:56

Someone was working real hard on this and then got distracted.

😂 1
martinklepsch13:11:46

Distraction Driven Open Source Development

martinklepsch13:11:17

(distraction has a bad connotation but you might as well say “curiosity and fun driven development”… it just won’t sound as cool 😎)

lread14:11:27

Some people are driven to distraction, but I just walk.

rkiouak23:11:44

hm, I could use some help debugging an analyzer error… I’ve reached out on this previously — the hypothesis was that the extra .class files of dependencies that hadn’t been removed when AOTing was causing this error https://app.circleci.com/pipelines/github/cljdoc/builder/25501/workflows/8ce22de5-1860-4409-95e8-9cc7cb0e0029/jobs/41876, but I’ve now resolved the issue with my library and the .jar contents now only reflect source. Any ideas what may be causing the “Can only recur from tail position” error only when under analysis?

lread03:11:40

Was this something to do with ClojureScript source failing to analyze?

rkiouak03:11:35

I believe the cljc issue will present after this tail recur error is resolved. There are two separate issues, IIUC. Previously, I had pulled the nses where this error were occurring into a separate lib, as I was unsure how to resolve the AOT’d .class file inclusion — now that I’ve fixed this issue in the jar construction and have reintroduced the nses, this earlier recur error is presenting again. Once we sort this, I can look more closely at the .cljc/cljs issue and resolve it next.

lread19:11:05

Ah… my brain has forgotten all details of this.

lread19:11:00

Yeah, still holds true. If I downgrade protojure to core.async 0.4.490, I’m past the recur error and now at your ClojureScript error.

lread20:11:10

I honestly don’t know, but am wondering if making your AOT classes thinner would be a good thing for protobuf in general. Just make them a proxy that delegates to Clojure maybe? Here’s https://clojure.org/reference/java_interop#_calling_clojure_from_java, and here’s https://github.com/clojure/tools.deps.alpha/blob/master/src/main/java/clojure/tools/deps/alpha/util/S3TransporterFactory.java.

rkiouak22:11:32

Ah, i dont know that i caught these details previously. Thank you for rehighlighting, ill take a look at your links -- we need to implement some interfaces and instantiate/reify an object of a particular class for java webserver interop.

lread02:11:56

Yeah, no problemo @U7F53SKPD, it has been a while!