print 'ok' after connection

[ofrobots]

This change makes inspect more compatible with the old debugger. There
are some Node.js core test-cases that are expecting the 'ok' to show up
here.

I am investigating if we can make node debug an alias of node inspect.

[jkrems]

Ran npm test locally - it fails because of commit message conventions (it expects a prefix like fix: ).

The change itself LGTM and the actual test suite passes. 👍

[ofrobots]

Commit message fixed.

[jkrems]

Verified npm test locally. Not sure what the official flow with node CI is. I'm guessing a member of the diagnostics WG would now start a CI run against this branch..?

[ofrobots]

Hmm.. let me see if I did this right.. https://ci.nodejs.org/view/x%20-%20Diagnostics/job/node-inspect-continuous-integration/6/.

/cc @mhdawson does @jkrems need to be a collaborator to get access to the CI?

[jkrems]

Looks like there's some inherent instability in the Windows build. Maybe caused by different stdout flushing behavior/timing..? Don't think that's connected to this PR though.

[mhdawson]

Anybody who is part of the diagnostics team should be able to kick of the node-inspect job. Looking at the list I don't see jkrems on that list but given his involvement in node-inspect my feeling is that he should be but I think we need members of that WG to chime in. @joshgav @nodejs/diagnostics

[mhdawson]

And specifically, he does not need to be a collaborator, just a member of the team configured for the diagnostics workgroup.

[mhdawson]

@jkrems I did open some issues for failures in the CI. My guess is that we'll find they are issues with the tests, most likely related to timing. May of the failures we've seen when broadening out the platform coverage tests run on fall into this category. Do you have cycles to look at those failures ?

[joshgav]

Yeah we need to manage membership of the Diag WG better, I've been waiting for the changes in the TSC to land though. PR to add Jan: nodejs/diagnostics#87

[mhdawson]

@jkrems at this point unless you see new failures in the CI jobs that were not in the previous runs without the commit, I think this can go ahead and land.

We'll need to work the failures to get to green on all platforms @CurryKitten is starting to look at the zLinux/AIX failures. @joshgav do you have anybody you can nominate to look at the windows failures ?