Re: [Rescert] making the "torn dollar bill" and the swing point (fwd)

[Andrei Robachevsky <andrei@ripe.net>]

George Michaelson wrote on 24-07-2007 17:40:
> really, having dived into the crypto thing in a very annoying way, I'd
> like to dive back out again and leave it as a problem statement. every
> time I try to use crypto language, I just make a fool of myself.
> 
> my take on the problem statement:
>

It is a very good outline George.

> 	under the control of the swingpoint
> 
> 	two parties, buyer/seller, from/to, whatever
> 
> 	*not always directly under swingpoint*
> 
> 	swingpoint needs to make something that retains its control
> 
> 	seller has to authoritatively commit to sell
> 
> 	buyer has to authoritatively commit to buy
> 
> 	completion is when the swingpoint gets told both have
> 	formally committed and that the from/to behaviour at the
> 	swingpoint should be done.
> 

What is the completion in this case? I understand that any changes in
RPKI start only when both have formally committed (e.g both parts of the
torn dollar bill are received"? And this process is irrevocable.

Andrei


> 	parties along chains to swingpoint between buyer and seller,
> 	separately have 'interrest' and need to flag consent or 
> 	understanding (is this really true?)
> 
> and a belief that something, in keypairs can help us. 
> 
> I think my M of N issue is probably not relevant.
> 
> I think my 'make new keypairs' thing is probably not relevant, but I
> still have doubts that business logic IDs are going to work.
> 
> -G
> _______________________________________________
> Rescert mailing list
> Rescert@apnic.net
> http://mailman.apnic.net/mailman/listinfo/rescert