Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: Testing...hello??? Helloooooo???

Switch port??

Why switch port?
Many of these devices will have low occupancy on the net (e.g. light
switches and thermostats) and will have no need of a switch port. Repeater
based hubs will do quite nicely thank you. Although the cost difference is
small in # of $ it is still large on a % basis at the retail end of the chain.


At 03:39 PM 3/20/00 -0800, Rick Brooks wrote:
>I'm not sure that I know what you mean by "single switch port" (or did you
mean switch box?).
>To me, there is one DTE device per switch port. And the proposed rule is:
only the switch port can put power on the cable, and only the DTE device
can take power off of the wire, right? Also, you cannot have multiple DTE
units that take power off of the same wire.
>I think that the reason to bias the low cost end of the wire to the DTE
end, is to allow the lowest cost and simplest types of products to be
developed for the DTE end. The per port cost at the switch end is also
important, but perhaps not as important. Obviously, it would be nice to
have an identity network that can eventually be completely integrated to
lower cost and board space.
>- Rick
>At 02:36 PM 03/20/2000 -0800, Norman Finn wrote:
>>Rick Brooks wrote:
>>> I wanted to add a few things to you list.
>>> 1) goal: achieve the lowest possible cost at the DTE load end:
>>>    a) possibly allow for non-isolated DC/DC converter (in those cases
where safety rules allow this)
>>>    b) simplest lowest cost identity network at the DTE load end
>>Depends on the application.  For an application for which there are many
>>DTEs for a single switch port, this might be true.  But for a great many
>>applications, e.g. a desktop IP telephone, this is not true.  The owner
>>must pay for the equipment at both ends of the wire.  Why is there a
>>particular need to bias the costs toward one end of the wire or the other?
>>(: Yes, I understand that the switch end typically is serving a number of
>>wires.  I'm talking about the per-port cost. :)
>>-- Norm
| Geoffrey O. Thompson                    |
| Chair IEEE 802.3                        |
| Nortel Networks, Inc.  M/S SC5-02       |
| 4401 Great America Parkway              |
| P. O. Box 58185                         |
| Santa Clara, CA 95052-8185  USA         |
| Phone: +1 408 495 1339                  |
| Fax:   +1 408 495 5615                  |
| E-Mail: gthompso@xxxxxxxxxxxxxxxxxx     |
| Please see the IEEE 802.3 web page at   |