I want to add to the chorus here: technically the code is owned by him and any other contributors to the module (unless he used a contributor agreement that waived their rights; there are also some legal grey areas around small contributions that don’t qualify as “works” under copyright law).

This is as opposed to public domain code, which really is technically owned by nobody.

What he waived was his right to tightly control distribution. By releasing under an open license, he lost the ability to say who his software was used by, and how it was distributed (outwith the terms of the open source license he released it under).

This is a crucial distinction — particularly when considering funding for open source! Here’s one example: when the author retains their original ownership rights, they are free to relicense the code, or simultaneously release it under a commercial license, for example with different support provisions. It also opens the code up to being part of an acquisition: when you have ownership, you can transfer ownership. Both of these are difficult to do if you have waived ownership.

(Paranoid disclaimer: I’m not a lawyer and this post doesn’t represent legal advice.)