This is kind of half feature request half bug report 😛
1.5.x introduced ‘setImmediate’ pollyfilling, which is great, but in our case it was kind of a regression because it meant when we updated, we had a new rewrite that polyfilled setImmediate calls that were previously not rewritten. That caused some of our tests started failing because of slight timing differences between the new setImmediate polyfill and
Q own internal polyfill.
I’ll just set
setImmediate: false now, but I’m now slightly concerned about updated in the future silently starting to inject polyfills that I don’t expect. Not necessarily the end of the world, but we try to track all 3rd-party code that we package and distribute, and this would have probably been missed if not for the test failures it caused.