Package Rename bug fix

[confused-Techie]

Requirements

• Filling out the template is required.

• All new code requires tests to ensure against regressions.

  • However, if your PR contains zero code changes, feel free to select the checkmark below to indicate so.

• Have you ran tests against this code?

• This PR contains zero code changes.

Description of the Change

As helpfully discovered by @savetheclocktower we currently, mistakenly check for the truthy return of db.packageNameAvailability to fail on a package rename, when we need to check it's falsey return.

[savetheclocktower]

I just observed the effects — you found the bug :)

[DeeDeeG]

I assume what this code means is, as of the fix, is that if new name is not available (!isAvailable.ok) we print an error, otherwise we (the backend) continue on to publishing a package and doing all that other stuff.

Seems legit! I can see why that being inverted before could count as a bug.

(Mostly piggy-backing off of confused and savetheclocktower saying this seems to be right, TBH. But this seems like a sensible fix to my own judgment as well. Just, I'm not that familiar with the backend code, is all.)

[DeeDeeG]

If it's alright with everybody, I'll wait on this to land before approving pulsar-edit/ppm#135, since as it is right now I can't end-to-end test that that ppm fix is working, whereas this one appears if I'm following conversation correctly to be the real root-cause issue, and the ppm one is a more superficial one (or at least a client bug rather than infra bug).

Maybe that rounds down to my having one reason: I'd like to be able to test better, heh.

[confused-Techie]

Thanks a ton for the reviews @savetheclocktower & @DeeDeeG I'll get this one merged and deployed