Policy-Discussion

[Ian G <iang AT cacert.org>]

Hi Alexander,

writing from the pov of the AP:

On 21/01/2010 12:54, Alexander Prinsier wrote:
> On 01/20/2010 10:42 PM, Pieter van Emmerik wrote:
> > The way I interpret the rules you have the freedom to record multiple
> > variations of the name as long as the can be verified
> > from a government issued ID.
>
> It's good you interpret them like this, but history has shown not
> everyone does it like this. Why not make it *clear that this* is the
> right way* to do it?


You might be missing a key point.  The AP is ahead of the system. 
Unfortunately, the system has not the mods in it to permit the multiple 
writings of names.  (fyi, we consciously held back this modification 
because it doesn't impact audit, and there are 3 other major 
audit-impacting mods to the system required, c.f., the infamous CCA 
rollout programme.)

It is somewhat futile to make the AP "clearer" about a situation that 
isn't possible.  It is far better to fix the system to do what we want 
to do, then people will see the system & the AP, and understand it better.


> I've heard arguments like "let's wait and see what an arbitrator rules".


That's not an argument, per se, it's a policy & strategy :-)


> Do we want the arbitrators to make policy then? It's the policy group's
> task to state clearly *what they mean with what they write*.


Ah.  That's missing a key part of CAcert's governance structure.  The 
way it works is this.

Policy group writes the policy, as best and /concisely/ as possible. 
Board implements it (or more probably, stands aside while teams do 
it...).  Arbitration deals with the exceptions, messes, contraditions, etc.

So, it is fully and consciously intended that Arbitration deals with the 
complication known as "what's your name?"  This was entirely and 
explicitly in the minds of the people who wrote the AP.  That's because 
they were all quite happily aware of the complications, having been 
around the planet and seen a lot of different cultures, done a lot of 
assurances, etc.

More precisely, the rules-based approach was deliberately and 
concsiously rejected by the writers because it is too hard to get right 
and to get tight.


> Currently there is no consensus about what has been written in the AP
> and practice on names. Some are in favor of a strict match with
> government ID, some are in favor to allow non-technical variations which
> are in no government ID, *even* in Europe.


Correct, or not, depending.  It was the AP team's consensus that there 
is no such thing possible as the correct way to write a name.  (And 
there is no such thing as a correct name, nor even one name.)


> Hence even the policy group doesn't know what they wrote, imagine what
> an arbitrator has to make of it.


<cough :>  The policy group knows literally what they wrote, it's in the 
policy :)  What might be lost in time is why they wrote it that way, but 
so far that hasn't happened!


> I'm only asking to *clarify what is allowed and what is not*, by
> amending the policies..


OK, I see what you are asking.  I'm pointing out, that wasn't intended. 
 We explicitly intended for:

    the Assurers to find and isolate those
    the Assurance Officer to consider and write a practice
    the Arbitrator to clarify each detailed dispute
    the Handbook to record that practice
    subpolicies to be developed were needed
    etc, etc

It was quite clear to the authors that there is a "European model" for 
names which is very strict in lettering and exactness and government ID, 
etc.

That strictness was not adopted by the writers of AP.  Quite 
consciously.  Instead, AP and other policies did *other things* which 
mesh nicely into this.


> > However the name can be recorded different on different ID like name
> > different on passport and drivers license.
> > As I see it transliteration is allowed as described in
> > http://wiki.cacert.org/PracticeOnNames.
>
> Transliteration is different from abbreviation for example.
> Transliteration often happens for technical reasons, which my proposal
> is OK with.
>
> > Translations from the name as it appears in an ID should not be allowed
> > as it can not be verified.
>
> What about an abbreviation which is not on any government issued ID,
> even though the abbreviation is very common? 'Alex' instead of
> 'Alexander', 'Paul' instead of 'Paulus'.


So, given the above, and recalling the mindview of the framers of AP, it 
was intended that this and other questions were to be resolved by 
Assurers, Assurance Officer, Arbitrators, Handbook and subpolicies as 
desired.


> The second is an example of an abbreviated Christian name. Some people
> are given Christian names at birth. Some of those don't want to use them
> afterwards and use 'Paul' instead of 'Paulus' for everything. Some of
> them don't have any ID that states 'Paul'. What's the policy group's
> opinion about that?
>
> > As worded in the proposal my vote is:
> > Naye
>
> Would you be in favor of a Europe subpolicy?


Technically, if there was a need to set up a policy distinction, this 
was how it was to happen.  Write a European subpolicy for Names.

(However, I'm writing that as an interpretation of the AP, not as a 
personal opinion as to whether this is a desirable thing or not.)


iang

> Would you please reword it
> so that you would agree?
>
> Alexander

Attachment: smime.p7s
Description: S/MIME Cryptographic Signature