On 2020.05.22 18:52, Thomas Schmitt wrote:
remote: error: denying non-fast-forward refs/heads/win32_first_tno_fix (you
should pull first)
Yeah, that error is typical of git repos where you can't use force to
push changes.
On one hand, not allowing force ensures that people who cloned the repo
don't run into issues if they picked commits that were discarded or
altered afterwards (though my understanding is that git is smart enough
to let you know and you can always use git reset and pull again).
On the other hand, force-pushing a commit allows you to fix mistakes and
keep things a bit cleaner.
I'd say you probably want to disable the ability to force push on a repo
that sees a lot of traffic or if you granted push access to more than a
few people (because you never know if someone's computer could be
compromised and used to alter existing changes, though one way to detect
that is to use signed commits, as one would also have to steal a private
key or password to make forced commits look legitimate).
I don't really have a strong opinion about whether force push should be
enabled for libcdio or not, as we can manage either way.
Thanks for doing the work of making compilable, testing, and committing.
(Actually i wonder what i did except spotting the problem and cluelessly
copying lines from gnu_linux.c.)
You're giving yourself way too little credit here!
The reason why you may not feel like you did much work is because you
were able to analyse the problem quickly and knew where to look. So
don't short-sell expertise, especially at a time where a lot of people
seem to think forming an uneducated opinion about a subject matter
counts as much as having real expertise on the same subject... ;)
Regards,
/Pete