[E-Lang] ERTP-aware MintMaker

Mark S. Miller markm@caplet.com
Wed, 14 Feb 2001 08:46:07 -0800

At 06:09 PM Tuesday 2/13/01, hal@finney.org wrote:
>Steve Jenson writes:
>>                 to compareTo(other) :any {  
>>                     quantity compareTo(other.quantity)
>>                 }  
>Maybe it would be a good idea to have the issuer vouch for "other"
>in compareTo?  I gather that it is supposed to be another assay.
>(I forget why you want to be able to compare the amounts in two assays.)
>Probably a fake assay would be detected later but in general it is a
>good idea to detect errors as early as possible.

In general I agree, and I believe I agree here as well.

As to "why compareTo()?", it is the most mysterious part of the protocol, 
and I don't think it's explained anywhere yet, so here goes:

The driving purpose of the protocol is to enable a 3rd party mutually 
trusted escrow exchange agent to engage in what might call "oblivious assay" 
and/or "oblivious escrow".  Look at the last diagram on 
http://www.erights.org/smart-contracts/exchange.html while reading the 
following.  Btw, the following description applies almost equally well to 
the SchedVid market within the WebMart marketplace built by Agorics & Sun 
Labs around 1996, using the ancestor of ERTP 
( http://www.agorics.com/webmart.html , 
http://www.sun.com/960201/cover/webmart.html , 
bottom of http://www.sun.com/research/techrep/1995/annualreport95/ist.html ).

Betty Bank is a money issuer.  Joe-Dish Future & Title is an issuer for 
"Satellite Dish Futures".  Such a future gives the right to, for example, 
determine where the Sun Labs rooftop satellite receiver dish points between 
2 and 3 tomorrow.  The first kind of good (money) is fungible.  The second 
is not.  However, both are divisible and poolable.  The above right might be 
divided into two separate rights: to the interval 2-2:20, and the interval 
2:20-3.  Given that such rights are held in a Purse in an ERTP(-like) 
protocol, rights from the same issuer must be generally combinable, so a 
Purse might actually hold rights to point during a sequence of disjoint 
future periods of time.

As we'd guess from the nature of Issuers, in order for Mr. Slippery to buy a 
Future from Alice, Betty and Joe must both have prior knowledge and trust in 
these two issuers.  Their knowledge of what they even *mean* by these 
instruments, and their basis for valuing them, can only be rooted in their 
prior relationship with these issuers.  In some sense, these issuer 
identities are the roots of what these instruments are.

Also, unsurprising, Betty & Joe don't need to have any prior knowledge of 
any of the other parties on the diagram.

In order to bring about an all or nothing mutually acceptable trade, Mr. 
Slippery and Alice need to also agree on the identity of an escrow exchange 
agent that will obtain into its own escrow the rights each is offering to 
the other, in order to ensure that both sides of the trade may happen, and 
that the rights involved are indeed mutually acceptable to the other.  If it 
fails to complete the trade, it will release any rights acquired back to the 
originating party.  Only once it knows it can complete the trade does it 
then release the obtained rights crosswise.

Now the part that may be novel.  (If someone's even articulated these goals 
before, I'd love to hear about it!)

By using ERTP to interact with the other parties, the Escrow Exchange agent 
can perform its function as well without any prior knowledge or trust of any 
of the other parties on the diagram.  Although it is assaying and escrowing 
rights, and calculating mutual acceptability, it doesn't know, and doesn't 
care, about the nature of the rights being transferred.

More later today...