You are not logged in.
In order to smooth out the stark difference in vetting quality between extra and aur packages, we could make use of the information about how many people (aur registered users) have checked the diffs for a certain package version. This would be a soft indicator of how many eyes have seen a certain diff (and found nothing worrying). This signal would encourage people to direct however much energy they have for checking the diffs into the ones that perhaps need the most.
Like votes it'd be self-reported and tool-assisted (via web would be too much friction). Unlike votes it'd be for versions of packages, not packages themselves.
I tried to find evidence of this being raised already. Disregard if it has.
Thoughts?
Offline
People that aren't checking diffs now will not check the diffs then, people that don't understand what kind of information they are looking for will look at the diffs and not be able to make a proper judgement call. People that are already properly checking diffs are checking diffs whether a helper feature tells them to or not and will bring potentially necessary reactions to the AUR comments. To me this feels like a bunch of engineering/computational overhead for something that will make 0 difference on how the relevant tooling that already exists will be used by people that already make use of it.
Last edited by V1del (2023-06-26 14:54:21)
Offline