test: fix too optimistic guess in setproctitle

[vsemozhetbyt]

Checklist

• make -j4 test (UNIX), or vcbuild test (Windows) passes
• commit message follows commit guidelines

Affected core subsystem(s)

test, process

test/parallel/test-setproctitle.js assumes that process.title can't be test before the test changes it. However, if this test is launched as a child several times, process.title is saved in the parent on Windows and is shared with childs.

I've found this issue running tests with some script with different options. A reduced example (edit the path to reproduce):

const { execSync } = require('child_process');
const cmd = 'node j:/temp/_git/node-fork/test/parallel/test-setproctitle.js';

console.log('run 1');
try { execSync(cmd); } catch (err) { console.log('run 1 error'); }
console.log('run 2');
try { execSync(cmd); } catch (err) { console.log('run 2 error'); }

Before this PR:

run 1
run 2

assert.js:86
  throw new assert.AssertionError({
  ^
AssertionError: 'test' !== 'test'
    at Object.<anonymous> (j:\temp\_git\node-fork\test\parallel\test-setproctitle.js:19:8)
    at Module._compile (module.js:582:30)
    at Object.Module._extensions..js (module.js:593:10)
    at Module.load (module.js:516:32)
    at tryModuleLoad (module.js:466:12)
    at Function.Module._load (module.js:458:3)
    at Function.Module.runMain (module.js:618:10)
    at startup (bootstrap_node.js:144:16)
    at bootstrap_node.js:548:3
run 2 error

After this PR:

run 1
run 2

It seems this better be fixed to prevent such issues with various test approaches to test base.

[vsemozhetbyt]

CI: https://ci.nodejs.org/job/node-test-pull-request/7800/

[Fishrock123]

AssertionError: 'test' !== 'test' ... err what is going there? Different encoding?

[lpinca]

@Fishrock123 https://github.com/nodejs/node/pull/12792/files#diff-97b9a9f4bc2e740f654ca4065d2e29fcL19

[vsemozhetbyt]

@Fishrock123 Actually, the 'test' === 'test' here, what is against assert's expectation.

[Fishrock123]

Ah sorry, my bad.

[jasnell]

That's confusing. Perhaps it's worth changing the message?

[vsemozhetbyt]

@jasnell Do you mean commit message? What would you advise?

[jasnell]

No, the message on the assert error

[vsemozhetbyt]

What message would you add?

[Trott]

That's the default message from assert.notStrictEqual(). I recommend leaving it exactly as is.

[vsemozhetbyt]

Landed in f971916

[gibfahn]

Is this okay to land on v6.x?

Should land with #13072