Hi,

On 28/10/2014 20:17, John Curran wrote:
> ... snip ...
>
> I'm certain there is a simple answer for this question, but it alludes me
> at the present time...
> 
> Given the risks of full resource list invalidation due to overclaiming, why
> aren't distinct certificates used for distinct resources?  If this is not 
> practical in general, wouldn't it at least be prudent to "groom" resources 
> that are going to be transferred into their own certificate so that the rest 
> of the resources held by the original child are not put at validation risk 
> (if a coordination error were to occur in subsequent transfer processing)

I believe you are right, in fact, I'm pretty sure, and I discussed this
with Mark a few weeks ago, that it's possible to get quite close to the
proposed semantics of validation-reconsidered just by separating
(grooming to use your term :-) ) resource sets accross different certs,
according to specific criteria.

This, to me, means that validation-reconsidered is actually is more like
an optimization proposal to something that could already be done today
iif we had HSMs able to support enough keys/certs. I'm convinced that
there isn't here a radical departure from the original spirit of
validation rules.

> 
> Thanks!
> /John
> 

cheers!

-Carlos
> 
> _______________________________________________
> sidr mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/sidr
> 

_______________________________________________
sidr mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/sidr

Reply via email to