isNodeJS check

[domarmstrong]

Hi, in asap.js you are checking whether the environment isNodeJS.

if (typeof process !== "undefined" && process.nextTick) {
    // Node.js before 0.9. Note that some fake-Node environments, like the
    // Mocha test runner, introduce a `process` global without a `nextTick`.
    isNodeJS = true;
}

We do have a fake environment with browserify where process.nextTick is used and this check is causing errors to be thrown synchronously causing issues.

One suggestion from ksmeltzer was to use process.title === 'node'. The problem is we have window in both envs and process etc..

[ghost]

Or Object.prototype.toString.call(process) == '[object process]'...

[kriskowal]

Do these environments also provide a bad process.nextTick?

[kriskowal]

Come back and re-open this if process.nextTick is bad on these fake Node environments. The isNode test just means that Q will use process.nextTick and check whether process.domain exists when necessary. It should not have any harmful side effects.

[domarmstrong]

The problem is this in flush:

           try {
                task();

            } catch (e) {
                if (isNodeJS) {
                    // In node, uncaught exceptions are considered fatal errors.
                    // Re-throw them synchronously to interrupt flushing!

                    // Ensure continuation if the uncaught exception is suppressed
                    // listening "uncaughtException" events (as domains does).
                    // Continue in next event to avoid tick recursion.
                    if (domain) {
                        domain.exit();
                    }
                    setTimeout(flush, 0);
                    if (domain) {
                        domain.enter();
                    }

                    throw e;

                } else {
                    // In browsers, uncaught exceptions are not fatal.
                    // Re-throw them asynchronously to avoid slow-downs.
                    setTimeout(function() {
                       throw e;
                    }, 0);
                }
            }

It thinks that the browser isNodeJS and is therefore throwing errors synchronously instead of async as it would otherwise and flush is not completing.

[slorber]

@kriskowal as mentionned in the linked issue I got a problem with this too.
In case of some error, Q completely stops working, producing important problems in production (SPA needing F5 refresh, as not a single promise can resolve anymore)

@dominataa is right, the problem seems to be related to the synchronous exception rethrow in case of isNodeJs = true, in case of a Browserify context.

This would be nice to be able to disable the "fail fast flushing behavior" used for NodeJS (In node, uncaught exceptions are considered fatal errors. Re-throw them synchronously to interrupt flushing!).

[slorber]

This problem appeared between releases Browserify 8.0.3 and 8.1.0. Q works fine with older versions of Browserify

[slorber]

I've made a PR for node-process/browserify that solves this problem.
See defunctzombie/node-process#37

[calvinmetcalf]

can confirm Q does some funky stuff in browserify because it thinks it's in node

[pho3nixf1re]

This happens with Webpack as well. It's killing our test suite and causing early exits instead of just test failures. It is definitely a bug with the flush in Q and not with the shim's in browserify and webpack.

[kriskowal]

This has been addressed by @kahnjw in v1.2.1. Thanks for the tips.