Subject: Re: mac <-> pc serial cable for OFW serial console?
To: None <port-macppc@netbsd.org>
From: S. Schwerzmann <sschwerzmann@mus.ch>
List: port-macppc
Date: 02/24/2003 21:14:37
hello,

> > So I've gotten a DIN8 serial cable connector.  Now my question is
> > this: do I need to use a null modem cable, or should the wiring
> > between the Thinkpad and the powermac be straight through?  The
> > following, from http://www.netbsd.org/Ports/macppc/faq.html ,
> > makes me think that the cable should be straight through:
>
> You need a null-model.
>
> > Your serial console settings should be 38400 bps, 8 bits, no parity,
> > 1 stop bit, no flow control, and you can typically use a standard
> > Mac "printer" cable to connect two Macs (even m68k-based Macs). See
> > the NetBSD Serial Port Primer for additional help.
> >
> > I guess that printer cables would be straight through.
>
> No. Mac-printer cables are null-modem cables.

   understanding the technology, part 194735e583g73.459a573.2658b5.tris

   Q:  where do the RS-232/RS-422 come from? what was their original 
purpose?
   A:  (once upon a time) they where to connect Data Communication 
Equipment (DCE, that is mostly _modems_ of any sort) to Data Terminal 
Equipment (DTE, that is computers, terminals, printers, etc.)
   the usual setup then looked like:

	DTE1---<RS*>---DCE............DCE---<RS*>---DTE2

   replace DTE1 with your favorite flavour of computer
   replace RS* with yout favorite flavour of serial line standard (stds? 
soo many to choose of...)
   replace DCE with your favorite flavour of modem
   replace DTE2 with your favorite flavour of peripheral (or other 
computer)

   and this usual case uses so called 'straight' wired cables, due to the 
fact that DCE and DTE connector pinout is symmetrically same and of 
different gender (if case shape allows, you can connect a DCE 'dongle 
style' directly to a DTE without a cable in between) which also means 
that, speaking in terms of direction, a signal on a pin comes _out_ of 
one device but goes _in_ to the other device on the same pin

   why was it thought like this? back then when standards where such, 
computers used to be room-filling noisy and heating monsters that 
usually where banned to the cave, but user terminals had to be on the 
desks of the users, obviously many floors (and thus many meters of 
cable) away - far too distant for 'simple' interfaces and modems hat to 
be put in for coping with the distance

   (I beg pardon for omitting some details, for the sake of simplicity...)

  - - -

   now what are many geeks trying to do and some struggle because of lack 
of understanding (or memory about the story above)?

   when 2 specimen of DTEs are close enough to each other, e.g. a 
computer and a printer  or  two computers (one disguised as terminal, by 
mean of running a terminal emulation program) 'just a cable' is more 
elegant than the cable/modem/cable/modem/cable combo
   one is tempted to omit the modems, on short distances - herefrom the 
term 'null modem cable', a cable that avoids the use of modems

   a closer look to the new  DTE---<RS*>---DTE  situation reveals that 
the connectors of the two devices are IDENTICAL, in terms of gender and 
also, on a given pin, the signal comes _out_ by both devices (or _in_ on 
both, on other pins) - obviously it makes no sense to bring two _out_ 
resp. two _in_ signals together
   ==> the null modem cannot be a straight one, the appropriate pairs of 
conductors must be crosswired so to bring signals from _out_ pins to 
_in_ pins

   find of pinouts for various cables at http://www.hardwarebook.net/

cheers
Stephan