Graham Klyne | 4 Aug 1998 12:38

Bid and asked

At 16:36 29/07/98 -0400, Al Gilman wrote:
>This sends me back to discussing a "bid/asked" distinction.  It
>may not be enough for each side to list tag/value pairs and for
>us to check for match on this pair of abstract vectors.

I think it is enough -- see later.

[...]
>If the demands asserted on each side are matched with compliance
>from the other side, then we have a feasible exchange, a match.

Yes.

>If we have more than one match, we look at differential stuff to
>achieve preference ordering.

Yes, but I see this as a separate issue.

>Is that already covered in the algebra as applications of an
>abstract match operation, or do we need to add some sort of
>mode, be it bid/asked or required/supported, to the tag/value
>entries to complete the logic?

My view is that the feature matching approach in -algebra- abstracts both
bid and asked scenarios into a common framework.  In actual *usage*, there
would offers of and requests for capabilities that must be matched, but
these are not distinguished within the feature matching framework.

#g

------------
Graham Klyne          _________
GK <at> ACM.ORG         ___|_o_o_o_|_¬
                   \____________/
(nb Helva)       ~~~~~~~~~~~~~~~~~~    <at> Stoke Bruerne, Grand Union Canal


Gmane