[Foundation-l] Fwd: Re: Fwd: Wikimedia-wide global blocking mechanism?

[Ray Saintonge]

effe iets anders wrote:
> I am talking here about *short term* blocks of proven vandal bots that
> are going through a high number of wiki's. These bots will not add
> anything good to the wikimedia wiki's, and are only destructive.
> Blocking here is stopping that bot from continuing, and I think that
> falls within the scope of what a steward is supposed to be for.
>   
This seems like shifting the goal posts.  As I understood the issue from 
the many messages in this thread it would apply to human editors as well 
as bots. How do you propose to distinguish whether a vandal is a human 
or a bot?  What is the process for establishing that it is a "proven" 
vandal bot?  How can you guarantee that the process will not be used 
against human editors who have merely been caught up in a pissing match 
with the admins in a particular project?
> The current situation is that as soon as a vandal bot is detected,
> there are two options. Either preventively block the IP on *every*
> small wiki, which means: go to the wiki, log in, grant yourself admin
> rights, block, remove the rights. A lot of work. Both result in long
> blocks, because we do not like the bots to come back tomorrow, it's
> too much of work for that. I think everyone agrees that this situation
> is bad and should be improved.
>   
If communities are going to insist that a steward desysop himself every 
time he goes into a project to perform routine anti-vandalism it's easy 
to see that it is more objectionable that they perform acts 
automatically without ever logging themselves in.  Perhaps it would be 
preferable that once a steward has properly given himself sysop rights 
he retain those rights unless there is a specific objection from the 
community.  That would be far more acceptable than any kind of automated 
process.

Ec