Anonymous contributions

[Ben Cooksley]

On Wed, Apr 17, 2019 at 9:08 AM Friedrich W. H. Kossebau
<kossebau at kde.org> wrote:
>
> Am Dienstag, 16. April 2019, 22:59:52 CEST schrieb Boudewijn Rempt:
> > I completely fail to understand what you're trying to say. This has nothing
> > to do with a commit hook that makes a misguided attempt at parsing strings
> > and validating them as real names.
>
> ? It has. For the start, it simply uses our existing big data of valid real
> names (a.k.a. identity.kde.org). For registered persons pushing, that would
> cover validating their name.
> And as nice bonus prevent unwanted accidental use of private data (like
> private/company email address or private user name).
>
> For unregistered persons, that validation would be shifted to the person doing
> the commit (!= author), who then would do the validation on their account.
>
> So this would actually implement what you ask for: no guess work about valid
> names.

Unfortunately validating against Identity data isn't a workable
option, because Identity most definitely requires separate first and
last names.
This isn't something we can easily change either, because we're bound
to LDAP schemas defined by various RFCs, which specify the attributes
"givenName" and "sn" as mandatory (being first and last name
respectively)

Removing this restriction is non-trivial, and would require us to
specify custom schemas and check that we aren't going to break any
application using Identity which is relying on every account having
both givenName and sn attributes.

The easiest way forward, as Nicolas mentioned, is to whitelist people
who have mononyms on a case-by-case basis.
Unfortunately for Identity they'll need to continue to have two names
provided until that system is replaced.

>
> Cheers
> Friedrich
>
>

Regards,
Ben