<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
</head>
<body bgcolor="#ffffff" text="#000000">
Our third residual RID is this one:<br>
<blockquote>
<p class="MsoPlainText">Sending a separate "cell_spec" MPDU for each
cell in the message space should be optimised to sending a single MPDU
that
contains a list of the
registrars for the cells, i.e. concatenate the "cell_spec"s.<br>
</p>
</blockquote>
<span style="font-size: 12pt; font-family: "Times New Roman";">My
proposed disposition is this:<br>
</span>
<blockquote><span
style="font-size: 12pt; font-family: "Times New Roman";">Not accepted.<span
style="">
</span>The proposed change would complicate the protocol by adding
another MPDU
format; currently, the cell_spec MPDU can be used in both registrar
registration and registrar location.<span style=""> </span>The
reduction in network traffic would be insignificant, and processing the
cell_spec MPDU would become appreciably more complex.</span><br>
<span style="font-size: 12pt; font-family: "Times New Roman";"></span></blockquote>
<span style="font-size: 12pt; font-family: "Times New Roman";">This is
actually quite separate from the node registration protocol changes
that we hammered out during the meeting: registrar registration is a
procedure that occurs very rarely and will involve a tiny number of
message exchanges unless the message space is truly mammoth, which I
very much doubt will happen in an on-board environment. I feel pretty
strongly that this is a false optimization, and I would really like to
set it aside.<br>
<br>
Scott<br>
</span>
</body>
</html>