Stuck on writing an end-to-end test helper. If I write "testcafe firefox ./tests/e2e" directly (in the command line) it works fine, but running that command through exec from child_process doesn't.

Here is my code: https://pastebin.com/RneJ4VGgI'm writing this script so I don't need to write two separate entries in package.json for testing with a simulated instance of Firefox and a simulated instance of Chrome, I just have one which runs ts-node ./tests/e2e/init.ts. As the title says, if I trigger the end-to-end tool (testcafé) directly, it works fine. But if I run this script then the simulated windows appear and the tests are carried out but there is no visual feedback from the tool. How do I retain that visual feedback, is there something else I need to kill in my code besides rl with .close()? Here is a screenshot of the error I receive instead of testcafé's standard feedback.

Submitted April 22, 2019 at 02:55PM by beefyjon