[g4] Re: [ca-tf] Draft pre-read document for the CA-TF workshop of 13 February

On 12.02 16:12, Leo Vegoda wrote:
> Hi Andrew,
> 
> On Feb 7, 2007, at 9:54 AM, Andrew de la Haye wrote:
> 
> [...]
> 
> >By the end of this week an additional document will be provided,  
> >containing open generic and very specific questions and discussion  
> >items.
> 
> Was this document sent? I can't see it in the mail archive on the web  
> site?

I am also curious about it. Unfortunately I cannot make the meeting today because
I am down with a serious coughing and sneezing thing. I believe part of this doc
was to be an exose around the following thoughts:

        - what is the service we propose ?
                - what is the service provided by us ?
                        - concrete service elements
                        - service description / draft documentation
                - how does one use it ?
                        - service description / draft documentation
                - change/new processes at RS
                - change/new processes at customer
                - ...

        - what are the benefits to the membership
                - potental simplification and strengthening of provisioning processes
                - potential for secure routing protocols
                - potential for secure transfer of number resources
                        - how will THAT work exactly?
                - ...

        - what are non benefits
                - miraculous improvement of registry data
                - replacement of registries
                - ...

        - what are the costs for the member
                - changed processes
                - cost of CA
                        - and *here* only start about the "CA service"

        - what are the costs at the NCC

        - what is the expected uptake
                - slow / fast ?
                - does uptake affect usefulness benefits ?
                - measures to promote uptake

 	- what are the consequences for members who do not take up the service?
 


I'd also like to offer the following to think about:

It is often implied that certification will improve the overall quality
of registration data and provide a better handle on who is the user of a
certain block of address space.  I argue that it is more likely that this
will not be the case:

1) New certificates for existing address space will be based on the
current registration data.  So by definition they cannot be more
accurate.

2) When certificates and registration databases co-exist both systems
will diverge and show different information. Is this an improvement?

3) When and if certificates supersede the registration databases for
operational purposes, the incentives to maintain the registration
databases will be reduced and registration databases will deteriorate.

Especially the last point deserves attention.  A lot of operational
coordination is based on the registration databases; will this still be
possible with degraded or seriously out-of-date registration databases?
Are we prepared to loose the capability of direct operational
coordination and to revert to a coordination model that follows peering
relationships only?

The registration databases also serve valid functions for
other users ranging from policy makers via law-enforcement to individual
Internet users. Deterioration of the databases will cause dissatisfaction
and resistance from those users. How are we going to deal with that?



Again apologies for not being there. Force majeure

cough/sneeze

Daniel