[Tfug] Harmless musing

[Bexley Hall]

Hi,

--- George Cohn <gwcohn at simplybits.net> wrote:

> Bexley Hall wrote:
> > 
> > So, what "little detail" am I missing that
> explains
> > why The Powers That Be opted to wire hubs/switches
> > one way and "computers" another?
> 
> Most commercial stuff is wired that way.  Think DCE
> - DTE and you can 
> understand why a straight through cable works.  The

Yes, but EIA232 isn't as "symmetric".  I.e. one can
understand why there is a difference between the
connectors (semantically as well as physically).

OTOH, network connectors -- regardless of whether
they are on a hub/switch or on a "terminal" ppiece
of kit actually resemble each other (logically -- with
just the pinouts differing)

> transmit on one end 
> connects to the receive on the other, etc.
> 
> With consumer grade stuff, I think they made them
> "universal" with a 
> circuit to detect which was xmit and which was rcv
> to cut down on 
> customer service calls which costs them money.
> 
> Why they don't do this on commercial class equipment
> is a mystery but I 
> suspect they figure if you paid big bucks for a
> switch, you know how to 
> wire them and they can save a few pennies on each
> port by not having the 
> auto detect circuitry.

But, my point is, why didn't they just *design*
them to be mirror images of each other from the start
and require *all* interconnect cables to have a
twist?  This seems *so* much smarter... ?

E.g., I can undersstand why a USB cable has A and B
connectors... based on it's *inteded* roles for the
two devices at each end of the cable.  But, RJ45
connections tend to be more like "peers"...

And, no apparent *electrical* reason for them to be
wired one way vs. another?

> At least this is the reason that was given in some
> of the Cisco classes 
> I've taken.  ;-)

<grin>


      ____________________________________________________________________________________
Get easy, one-click access to your favorites. 
Make Yahoo! your homepage.
http://www.yahoo.com/r/hs