[Ietf-not43] -02 requirements draft

[Rick Wesson]

On Tue, 5 Nov 2002, Andrew Newton wrote:

> So...
>  1) You raise an issue.
>  2) Conversation takes place
>  3) Language is agreed upon
>  4) You never objected to it

I've objected several times, in person and on the list.

> Given that you seem to be unwilling to work with us, then perhaps you should
> explain why it will not work.  And allegations of absurdity and allusions to
> black magic are generalizations.

ok, for all participitants in the mesh to answer a query of "show me all
the domains owned by john smith" all participiants must answer for a
complete result set at the client. My problem with this query isn't the
size of the result set is that ALL participitants must participate.

If there are 150 nodes in the mesh all 150 must answer. If 10 of the 150
node have 90% of all the domains 140 nodes answer negatively 90% of the
time. thus you have the burdon of the largest nodes imposed on the
smallest (in size of registrations)

This appears to be a big problem for small participitants in the mesh as
they will require significant bandwith and cpu just to answer negatively
for 90% of all queries they recieve.

90% of the participitants answering negatively 90% of the time is not an
acceptable solution for me. now if somone can show me how this isn't the
case for "3.2.3 Domain Registrant Search" I'll shut up.

-rick