> Editor's note on April 11, 2018: NPR has retracted the story that was previously on this page because it did not meet our standards. "Fairness" is one of our guiding principles, and to that end we have pledged to "make every effort to gather responses from those who are the subjects of criticism." In this instance, that did not happen.

So how does this not count as owning up to mistakes and accepting responsibility?

Your analogy with a git commit/`push --force` is flawed. In software, the end user is not at risk (without much deliberative effort) of running the buggy commit. In journalism, the article content is the actual end user product, and grievously erroneous content causes actual harm. Moreover, the mistakes that NPR admits to making were not just a "buggy" commit. It was the equivalent of checking in malware from the very first commit.