On Thu, Apr 21, 2022 at 12:53 AM Samuel Klein <meta.sj(a)gmail.com> wrote:
+++. We are raising these barriers to [apparently] try to stave off
vandalism and spam. But hard security like this can put an end to the
projects, for good. There is no more definitive end than one that seems
mandated from within. We need better automation, MLl models, sandboxing,
and triage to help us *increase* the number of people who can edit, and
can propose edits to protected pages, while decreasing the amount of
vandalism and spam that is visible to the world.
For the P2P proxy blocks, vandalism was a factor (AFAIK spam wasn't), but I
think the strongest trigger was the amount of harassment, death threats or
other physical harm threats, and doxxing attempts coming out from this
particular proxy service.
I agree that we should increase the number of people who can edit. But we
should also maintain a reasonably safe space for contributors. There are
trade-offs that need to tune at every corner.
For this kind of abuse, we have a toolbox:
- IP blocking
- Page protections
- Edit filters
- Bots and other post-edit analysis tools
- Manual patrolling (assisted with various tools) + reporting to
admins/oversighters/stewards.
Each of them has its own caveats, we should improve them all, and find some
balance in the usage of each tool. IMHO, complete removal of any of these
tools will be harmful to our projects and contributors.
Best,
Mario