[Gb-ccb] 4pm telecon

[Martin Shepherd]

Hi John,

On Wed, 5 Oct 2005, John Ford wrote:
>...The beauty of the original idea is that the IP
> address remains fixed for a CCB connected to a specific receiver.  In the
> new scheme, we'll have to call out computing when we replace a CPU module
> or complete CCB.  That's not really what we want to do, I think.

Agreed. This also came up during the telecon. Although the consensus
decision at the meeting was to leave things as they are for now, I
don't think that anybody felt strongly against going back and
implementing the original scheme. So if everybody agrees, then I will
implement this?

Note that it should be pretty easy to do. I already have a program
that will query the ID from the GPIO driver and display it, ready for
use in a script. I would need a list of fixed IP addresses to assign
to the each of the IDs. We would need one address for the 1cm
receiver, one for the 3mm receiver, and one for the lab harness.

When no cable is connected, the ID comes out as 00 (binary). We should
also use this ID for the lab cable harness, so that we can test the
CCB in the lab with or without a cable. I would also assume this ID if
the GPIO card failed or wasn't present.

Thus I suggest the following assignment of ID bits:

Binary ID    Cable harness
---------    -------------
    00        The lab cable
    01        1cm receiver.
    10        3mm receiver.
    11        A spare cable

Martin