[Foundation-l] (no subject)

[Gregory Maxwell]

On Tue, Oct 27, 2009 at 9:27 AM, Mike.lifeguard
<mike.lifeguard at gmail.com> wrote:
> Granted, you may argue that forcing requests to be made public would
> force more stringent compliance with the standards, but you haven't even
> shown that there is an issue in the first place; you have merely assumed
> that it is so. Beginning an argument with a false premise is generally
[snip]


Checkuser can cause harm in two primary ways:

(1) The wrongful release of private information about an editor. (IP,
Useragent, etc, or information derived from this private information
"Mike is in Chicago").
(2) Besmirching the reputation of good users; generally causing drama.

The written CU policy deals mostly with setting policy to prevent type 1 harm.

For type 1 harm to happen it requires the checkuser both perform the
check *and* blab about the results.  For type 1 harm to be
/meaningful/ the information had to be actually confidential: You
can't really say you were harmed by a CU disclosing your IP if you'd
been intermittently editing logged out with it for the past six months
then logging back in and fixing your signatures.

I think all checkusers take preventing type 1 harm seriously, even
though it is usually trivial for a motivated non-checkuser to obtain
an editor's IP, since the harm requires both a check and a disclosure
and we have reasonable policy covering it I don't expect type 1 harm
to be the most frequent problem.

Type 2 harm is magnified by the uncertainty of checkuser:  CU can't
really vindicate with any confidence, _at best_ it can really only
convict with a degree of confidence.  It has been a common sport for
some to accuse their enemies of sock-puppetry and clamouring for
someone to CU them. When the CU comes back with the inevitable
"nothing conclusive" that is used to sew additional doubt about the
accused.  Just the knowledge that someone saw fit to check is hurtful
to a person and their credibility.

So, conducting checks in private is the best tool we have for
minimizing type 2 harm: If no one but the CU knows that a check was
performed, no feelings are hurt, no credibility is smudged. If CUs
aggressively "turn down" requests made out of anger, even while
privately performing checks in the more suspicious cases, then drama
is reduced. The privacy in the checking process does create more
opportunities for type 1 harm, but as Mike mentions, I've not seen any
evidence that type 1 harm has been much of an issue in practice.