Fork me on GitHub

@juhoteperi: I tried your current patch and don’t get any errors when including cljs.js 😉


@juhoteperi: One thing I am not sure about though is when it should recompile dependents. I updated my small test repo and included a small example similar to the one you provided in the ticket ( When I change or cljs-1437-test.foo2 and then re-build, I would expect that cljs-1437-test.core is also recompiled, but this doesn’t seem to be the case for me. I am probably missing something in my setup?


@maria: Did you apply the patch? The cljs build number with the patch is 161 for me.


@juhoteperi: Should it also work with the build script, e.g. build once, make some changes and then build again?


@maria: I think it should work, but I can reproduce the problem.


I probably found the reason: ns-dependets doesn't correctly return dependents of namespace being compiled:


Could be because *inputs* is not bound with the new build pipeline


I'll try to fix this later today


Not sure why it works with watch though


Hey, we are trying to use the cljs testing system with phantomjs. We would like to exit nonzero if any tests fail. The recommended approach:

(defmethod report [:cljs.test/default :end-run-tests] [m]
  (println "\nYO!!!Ran" (:test m) "tests containing"
           (+ (:pass m) (:fail m) (:error m)) "assertions.")
  (println (:fail m) "failures," (:error m) "errors.")
  (aset js/window "test-failures" (+ (:fail m) (:error m))))
fails if there are any async tests. In fact, it isn't clear how to get any results when there are async tests.


basically I believe that :end-run-tests is not called if there are asynchronous tests.


Hi @chrisn we are using :end-run-tests for async tests successfully in doo


If I had to guess, your problem lies in the phantom runner, which is tricky to implement. I think this question is better suited for #C03S1L9DN


OK, good info. I will repost there and consider just using doo.