RE: [Geopriv] Resolving a design philosophy tension

From: Ted Hardie ^lt;>
Date: Wed Aug 15 2007 - 00:30:58 EDT

At 4:40 PM -0500 8/14/07, Winterbottom, James wrote:
>Ted, how do you see we proceed from here?

For the current issue, I believe the best way forward is to provide an optional
parameter for time the client is willing to wait, an optional parameter for accuracy
that the client desires, and to do the actual work of working how to relate the two.

I understand the points made by Doug and Martin in saying that accuracy
has always in their experience implied an upper bound. That doesn't
worry me much, since it will be entirely possible to express such an
upper bound with both available, and when an application arises
that doesn't want such a bound it will have the protocol mechanisms
available to make it work.

On the bigger point, I remain worried by the amount of talking past each other
we're able to do here, but I hope that Brian and Martin's agreement
that they both are aiming for "best basis" will help. If we can easily
build a broader agreement to that effect, then I think we have some
hope of eventually resolving the design conflict in favor of a useful
middle ground ("best basis with benefits?") . But there has to be a
willingness to examine our own assumptions and language here.
Reading through the past two thousand lines of fresh text sent to this
mailing list indicates to me we've got a ways to go.


Geopriv mailing list
Received on Tue, 14 Aug 2007 21:30:58 -0700

This archive was generated by hypermail 2.1.8 : Wed Aug 15 2007 - 00:32:38 EDT